2014 AGTA Conference Handbook

HANDBOOK
2014 AGTA
Conference
2
Handbook
2014 AGTA
Conference
CONTENTS
WELCOME 5
GENERAL INFORMATION 8
AGTA14 CONFERENCE PROGRAM 11
CONFERENCE SOCIAL PROGRAM 20
Abstracts and Biographies 21
POSTER PRESENTATIONS – Monday 59
POSTER PRESENTATIONS – Tuesday 62
SPONSORS 119
EXHIBITORS 121
EXHIBITION FLOOR PLAN 129
AGTA 2014 DELEGATE LIST 130
3
Handbook
2014 AGTA
Conference
AGTA
Executive Team
Dr Ruby C Y Lin (President), Asbestos Diseases

Research Institute, Sydney
Dr Carsten Kulheim, Research School of AGTA 2014 Conference
Biology, ANU, Canberra Organising Committee
Mr Mark van der Hoek, ACRF Cancer
Genomics Facility, Adelaide
Dr Alicia Oshlack, Murdoch Children’s
Ms Vikki Marshall, Genomics Core Facility,
Research Institute, Melbourne (Co-convenor)
Murdoch Children’s Research Institute,
Melbourne Dr Richard Tothill, Peter MacCallum Cancer
Centre, Melbourne (Co-convenor)
Dr Alicia Oshlack, Murdoch Children’s
Research Institute, Melbourne Professor Gordon Smyth, The Walter and Eliza
Hall Institute of Medical Research, Melbourne
Dr Richard Tothill, Research Division Cancer
Therapeutics Program, Peter MacCallum Dr Torsten Seemann, Victorian Bioinformatics
Cancer Centre, Melbourne Consortium, Monash University, Melbourne
Professor Erik (Rik) Thompson, School of Ms Vivien Vasic, MHTP Medical Genomics
Biomedical Sciences and Queensland Institute Facility, Monash Health Translation Precinct,
of Technology, Brisbane Melbourne
Dr Jac Charlesworth, Menzies Research Dr Mark Waltham, St Vincent’s Institute of

Institute, Hobart Medical Research, Melbourne
Associate Professor Daniel Catchpoole, Dr Noel Cogan, Department of Environment

Biospecimens Research and Tumour Bank, The and Primary Industries, Melbourne
Children’s Hospital at Westmead, Sydney Dr Kaylene Simpson, Peter MacCallum Cancer
Professor Ryan Lister, The University of Western Centre, Melbourne
Australia, Perth Professor Emma Whitelaw, Queensland Institute
Professor Ian Paulsen, Department of Chemistry of Medical Research, Brisbane
and Biomolecular Sciences, Macquarie Associate Professor Christine Wells, (AIBN), The
University, Sydney University of Queensland, Brisbane, and Institute
Mr Liam Williams, Centre for Genomics, of Infection, Immunity and Inflammation,
Proteomics & Metabolomics, Auckland University of Glasgow, Scotland
University, New Zealand Dr Andrew Fellowes, Peter MacCallum Cancer
Dr Mark Waltham, St Vincent’s Institute of Centre, Melbourne
Medical Research and Department of Surgery,
University of Melbourne, Melbourne
Associate Professor Marcel Dinger, Centre for
Clinical Genomics, Garvan Institute of Medical
Research, Sydney
Dr Kirby Siemering, Australian Genome
Research Facility (AGRF), Melbourne
Dr Mark Crowe, QFAB Bioinformatics, Brisbane
4
Handbook
2014 AGTA
Conference
ARC to consult with us for all things ‘omics”.

This year’s theme is aimed at enhancing
WELCOME FROM our understanding on clinical sequencing
THE AGTA with a specifically designed workshop. We
have focused themes on cancer genomics,
PRESIDENT epigenomics, plant genomics, bioinformatics,
transcriptomics, metagenomics and functional
genomics. On behalf of the managing
Dear Delegates and
executives and the local conference organising
Invited Guests,
committee, I would like to welcome our
On behalf of the Australasian Genomic international plenary speakers to Melbourne
Technologies Association managing and thank them for coming such a long way
executives, I welcome you to the 14th Annual and taking time out of their busy schedules, and
Scientific Meeting for AGTA in the heart of for their contributions throughout the meeting.
cultural Melbourne, Australia. We hope you enjoy and take time to explore
This meeting marks the beginning of AGTA, our beautiful country. We are also thankful to
which transformed from AMATA to reflect the our keynote speakers and national chairs. AGTA
evolution of genomics in Australia. When you 2014 committee is very proud to showcase
associate with AGTA you connect with many of many early career researchers (ECRs) and
Australia’s leading scientists. AGTA facilitates students. For the newcomers to the Society,
not only rigorous discussion on genomics but we trust that you will support our ECRs and
innovations, biology, translational research and students as they compete for Young Investigator
wet and dry laboratory techniques, across any Awards. The future of AGTA rests with our
species and various diseases. It is the premium younger members. Australian scientists continue
platform where people from academia, to capitalise on integrated technologies to
industry, research and clinic converge to tackle a diverse array of medical, agricultural
discuss major current topics impacting the and environmental, biodiversity and biosecurity
‘omics’ sectors. For the newcomers to the questions. I would like to thank each one of you
Association, this is where you form long-term for contributing to this dynamic line-up and I
friendships, network and collaborations and hope that you enjoy the program ahead.
find job opportunities. We are very thankful for the participation of
Our society marked a transformation in the our Gold Sponsor and Clinical Sequencing
past year with the introduction of the Small Workshop sponors Illumina, and our Gold
Grant Scheme that resulted in four successful sponsor Millennium Science incorporating
recipients, announced in January 2014. Two Fluidigm and Pacific BioSciences. We are also
of them will be presenting their respective thankful for our Bronze plenary session sponsors
posters at AGTA 2014 and two recipients will Bio-strategy, and some 19 or more exhibitors,
be announcing their respective bioinformatics without whom this conference would not
workshops in the coming months. We face be possible. We recognise the importance
lifted our website, agtagenomics.org.au in technology companies play in the evolution
collaboration with QFAB for hosting and ABN of genomics in Australia and you are strongly
for Yammer, and we have a social presence encouraged to take the time and opportunity
on Facebook, Twitter and LinkedIn. In to interact with these leading companies.
addition, our working groups have utilised It is my pleasure to welcome you to AGTA 2014
this momentum to generate awareness of on behalf of the AGTA managing executives
our society’s benefits amongst the student and the AGTA 2014 conference organising
populations. Furthermore, we also contributed committee. We are confident that you will
to the McKeon Review under the leadership enjoy AGTA 2014.
of ex-President A/Prof Daniel Catchpoole, and
Dr Ruby CY Lin
we are working towards having NHMRC and
President
Australasian Genomic Technologies Association
5
Handbook
2014 AGTA
Conference
WELCOME FROM THE

CONFERENCE CONVENORS
conference logo.
The AGTA conference enables the
opportunity for interaction between biologists,
bioinformaticians and technologists. This
unique mix is one of the reasons that the
Australian genomics community has a dynamic
cross-disciplinary and innovative approach
It is our great pleasure to welcome you to the
to genomic analysis, and is at the forefront of
14th annual conference of the Australasian
analysis tools for new types of ‘omics data. We
Genomic Technologies Association (AGTA)
have invited an outstanding list of international
at the Crown Promenade Hotel, Melbourne,
and national speakers plus have engaged
Australia, Sunday 12th to Wednesday 15th
industry to showcase the best in new genomic
October 2014.
technologies. We hope you make the most of
The AGTA conference (formerly known as the opportunity for some great networking and
AMATA) is Australia’s foremost genomic cutting edge science in the heart of the “most
technology conference. This year we have livable city” in the world.
worked hard to bring you an exciting program
We are excited to be holding the conference
that spans a diversity of research topics utilising
in Melbourne, one of the centres of Australian
cutting edge genomic technologies. Major
genomics research. We hope you enjoy the
themes include clinical genomics, functional
conference!
genomics, bioinformatics, metagenomics,
cancer genomics, epigenomics and plant
genomics. We have also scheduled a Alicia Oshlack
workshop focusing on the application of Murdoch Children’s Research Institute,
clinical genomics, an exciting area currently Melbourne
undergoing a major expansion in Australia.
Richard Tothill
Further features of our program include new
Peter MacCallum Cancer Centre, Melbourne
student social functions and our regular
conference dinner to be held at the Arts
Centre of Melbourne under the iconic
spire that has featured as part of this year’s
CONFERENCE MANAGERS
Leishman Associates
113 Harrington Street, Hobart TAS 7000
170 Elgin Street, Carlton VIC 3053
P. 03 6234 7844 F. 03 6234 5958
E. brigitte@leishman-associates.com.au
W. www.leishman-associates.com.au
6
Handbook
THE NEW CyTOF®2 MASS CYTOMETER
Discover more. Imagine more.
Mass Cytometry. The Future of Cytometry Today
Mass cytometry resolves multiple metal probes per cell with minimal signal
overlap, maximizing the information obtained from a single sample. With over
100 detector channels and over 30 isotopic probes, mass cytometry provides
researchers with an unparalleled ability to phenotypically and functionally
profile cells from normal and diseased states.
The DVS line of reagents, instrumentation, and data analysis tools provide all
the tools necessary for novel discovery applications using mass cytometry,
including:
VISIT THE MILLENNIUM SCIENCE BOOTH TO LEARN MORE
®
www.mscience.com.au
sales@mscience.com.au • 1800 678 242
2014 AGTA
Conference
GENERAL
INFORMATION
Registration Desk Conference Proceedings

The Registration Desk will be located on PowerPoints and abstracts will be available on
level one, Promenade Foyer Central. Please the AGTA website following the conclusion of
direct any questions you may have regarding the conference. Speakers will be requested to
registration, accommodation or social sign a release form. This is not compulsory.
functions to Leishman Associates staff at this
desk.
Conference Wi-Fi
Registration Desk Opening Times Wireless internet will be available throughout
the conference venue for the duration of the
Sunday 12 October 2014
conference. To gain access to the internet,
2.00pm – 6.00pm
connect to 'Crown Events and Conferences'
Monday 13 October 2014 wireless network. When you begin browsing
7.30am – 5.15pm the internet you will be prompted for the login
Tuesday 14 October 2014 details below. Please note that movies, music
8.00am – 5.30pm or illicit downloads are prohibited.
Wednesday 15 October 2014 Username: AGTA1

8.30am – 3.30pm Password: WrM20p
Join the conversation at #AGTA14.
Accommodation
If you have any queries relating to your Dress Codes
accommodation booking first speak to the Dress throughout the day is smart casual or
staff at your hotel or alternatively Leishman informal business.
Associates staff at the Registration Desk.
Your credit card details were supplied to the Emergency Medical Care
hotel you have selected, as security for your
For any medical emergency please telephone
booking. If you have arrived 24 hours later
000. The staff at your hotel will have information
than your indicated arrival day you may find
if you require contact details for a doctor,
that you have been charged a fee. You will
dentist or other health professional.
be responsible for all room and incidental
charges on check out and may be asked for
an impression of your credit card for security Social Program Entry
against these charges. This is standard policy in The Welcome Reception is included in the
many hotels. cost of each full conference registration.
The Conference Dinner IS NOT included in
Conference Name Badges any registration type. Social events ARE NOT
included in the cost of day registrations or
All delegates, speakers, sponsors and exhibitors
for accompanying partners. Places for day
will be provided with a name badge, which
registrants and additional guests for these
must be worn at all times within the conference
events may still be available at an additional
venue, as it is required for access to all the
cost. Bookings can be made at the Registration
conference sessions and social functions.
Desk subject to availability.
8
Handbook
2014 AGTA
Conference
All delegates who are registered to attend For information on the chairperson attending
the dinner will receive a named sticker at your session, please see the Registration Desk.
registration. You MUST place your sticker on A technician will be present in the speaker’s
a table located on poster boards next to the Preparation Room during registration hours.
Registration Desk. You must allocate yourself There will be facility to test and modify your
to a table no later than 11.00 am Monday 13 presentation as required.
October 2014.
Oral Presentations
Student Functions
Please refer to the program for the time
All conference students and early career allocated for each presentation, as these do
researchers are invited to the meet-a-mentor vary. The chairperson for your session will give
session and casual function on Monday 13 you a 3 minute warning, however you are asked
October, however registration is required as to adhere to your time allocation so that the
spaces are limited. Please forward enquiries program remains on schedule.
to the Registration Desk where bookings
can be made subject to availability. Further
information about these events can be found
Poster Presentations
on pages 20. Posters will be displayed in the Trade Exhibition
Proudly supported by: Area for the duration of the conference. Poster
Session One will be held on Monday 13 October
2014 from 1.30pm to 3.15pm. Poster Session Two
will be held on Tuesday 14 October 2014 from
1.00pm to 3.00pm. Further information on the
Poster Sessions can be found on pages 58.
Photographs, Videos, Recording

of Sessions Special Diets
All catering venues have been advised of any
Delegates are not permitted to use any type
special diet preferences you have indicated on
of camera or recording device at any of the
your registration form. Please identify yourself
sessions unless written permission has been
to venue staff as they come to serve you and
obtained from the relevant speaker.
they will be pleased to provide you with all pre-
ordered food. For day catering, there may be a
Speakers and Speakers' specific area for special diet food, please check
Preparation Room with catering or conference staff.
All speakers should present themselves to the
Speakers' Preparation Room, located in Room Disclaimer
M1, at least 4 hours before their scheduled
The 2014 AGTA Conference reserves the right to
presentation time, to upload their presentation.
amend or alter any advertised details relating
Speakers are requested to assemble in to dates, program and speakers if necessary,
their session room 5 minutes before the without notice, as a result of circumstances
commencement of their session, to meet with beyond their control. All attempts have been
their session chair and to familiarise themselves made to keep any changes to an absolute
with the room and the audio visual equipment. minimum.
9
Handbook
Historical Breakthroughs
Breathtaking Progress
Radical Reinvention
Introducing NextSeq™
A Whole Human Genome on Your Desktop
© 2014 Illumina, Inc. All rights reserved. Images

courtesy of Nature Publishing Group and Corbis Images. www.illumina.com/nextseq
2014 AGTA
Conference
SUNDAY
12 October 2014
1400 – 1800 Registration Open Promenade

Registration Desk
OPENING ORATION Promenade 1
Chairs: Dr Ruby CY Lin

1700 – 1800 Guest Speaker
THE HUMAN GENOME AS THE ZIP FILE EXTRAORDINAIRE
Professor John Mattick AO FAA
Garvan Institute of Medical Research, NSW
1800 – 2000 Welcome Reception & Trade Exhibition Proudly sponsored by: Promenade Foyer &
Promenade 2 & 3
Don’t forget
to get your Exhibitor
Passport stamped
to go into the draw
and win some great
prizes – including
an ipad mini!
KEYNOTE SPEAKER INVITED SPEAKER STUDENT EARLY CAREER RESEARCHER SPONSORED

11
2014 AGTA
Conference
MONDAY
13 October 2014
0730 – 1715 Registration Desk Open Promenade Registration

Desk
0800 – 1715 Trade Exhibition Open Promenade Foyer &

Promenade 2 & 3
0830 – 0845 Official Welcome and Conference Opening Promenade 1
SESSION 1: PLANT GENOMICS Promenade 1
Chairs: Dr Carsten Kulheim and Dr Noel Cogan

0845 – 0930 POPULATION GENOMICS UNRAVELS GENETIC DIVERSITY
AND REGULATION OF GROWTH AND DEVELOPMENT IN EUCALYPTUS
Professor Zander Myburg
University of Pretoria, South Africa
0930 – 1000 PLANT GENOMICS FOR CLIMATE ADAPTION
Associate Professor Justin Borevitz
Research School of Biology, ANU College of Medicine, ACT
1000 – 1015 Using small RNA sequencing to confirm and discover binding sites
of pentatricopeptide repeat proteins
Dr Katharine Howell
ARC CoE In Plant Energy Biology, University of Western Australia, WA
1015 – 1030 Discovering the Molecular Basis of Resistance in Australian
Myrtaceae to the Exotic Fungal Pathogen Myrtle Rust (Puccinia psidii
Ms Ji-Fan (Sarah) Hsieh
The Australian National University, ACT
1030 – 1045 Salt-specific responsive genes are revealed by transcriptional
profiling of Dunaliella salina cells in response to a reciprocal
change of salinity
Professor Jianhua Liu
Institute For Comprehensive Utilisation of Marine Biological Resources,
Zhejiang, China
1045 – 1115 Morning Refreshments & Trade Exhibition Promenade Foyer &
Promenade 2 & 3
SESSION 2: EPIGENETICS Promenade 1
Chairs: Professor Emma Whitelaw and Dr David Martino

1115 – 1200 3-Dimensional Genome Organization and Transcriptional Control
Professor Melissa Fullwood
Yale-NUS College, Singapore
1200 – 1230 shRNA screens for novel epigenetic modifiers
and characterisation of their roles in X inactivation
Dr Marnie Blewitt
Walter and Eliza Hall Institute, VIC
1230 – 1245 Characterisation of the novel epigenetic modifier Rlf
Dr Harald Oey
La Trobe University, VIC

12
2014 AGTA
Conference
MONDAY
13 October 2014
1245 – 1300 Widespread epigenomic and gender-specific

differences in isogenic mice
Ms Helen McCormick
Victor Chang Cardiac Research Institute, NSW
1300 – 1415 Lunch & Trade Exhibition Promenade Foyer &
Promenade 2 & 3
1330 – 1515 Poster Session 1 Promenade Foyer &

Promenade 2 & 3
SESSION 3: BIOINFORMATICS Promenade 1
Chairs: Professor Gordon Smyth and Dr David Goode

1515 – 1600 WHAT’S AHEAD FOR BIOLOGY?
THE DATA-INTENSIVE FUTURE
Associate Professor Titus Brown
Michigan State University, USA
1600 – 1630 The STATegra project: new statistical tools for analysis
and integration of diverse omics data
Dr Ana Conesa
Prince Felipe Research Center, Spain
1630 – 1645 Classification of genomic fusions into structural variation events
Dr Jan Schroeder
Walter And Eliza Hall Institute, VIC
1645 – 1700 Identifying fusion genes and other transcriptional variation in
cancer through de novo transcriptome assembly
Dr Nadia Davidson
Murdoch Childrens Research Institute, VIC
1700 – 1715 Developing an NGS pipeline for diagnostics
Mr Liam McIntire
SA Pathology, SA
1715 – 1815 Meet-a-Mentor Session
We invite students and early career researchers to join the
conference speakers for scientific speed dating.
1830 COMBINE Social Event
onwards PJ O’Brien’s, Southbank
Students and early career researchers are invited to join COMBINE for
a fun night of genomics trivia.
Walk with us from the conference venue at 18:15.

13
2014 AGTA
Conference
TUESDAY
14 October 2014
0800 – 1730 Registration Desk Open & Welcome Refreshments Promenade

Registraion Desk

Promenade 2 & 3
0825 – 0830 Welcome to Day Two Promenade 1
Proudly sponsored by:
Promenade 1
SESSION 4: CANCER GENOMICS
Chairs: Dr Mark Waltham and Dr Ian Majewski
0830 – 0915 Interrogating the architecture of cancer genomes
Dr Peter Campbell
Wellcome Trust Sanger Institute, Cambridge, UK
0915 – 0945 THE GENOMIC LANDSCAPE OF PRIMARY AND ACQUIRED RESISTANT
HIGH GRADE SEROUS OVARIAN CANCER
Professor David Bowtell
Peter MacCallum Cancer Centre, VIC
0945 – 1015 CIRCULATING TUMOUR DNA AS A LIQUID BIOPSY IN CANCER
Dr Sarah-Jane Dawson
1015 – 1030 A novel long noncoding RNA, lncUSMycN, promotes tumourigenesis
by binding to the RNA-binding protein NonO and up-regulating MYCN
oncogene expression
Dr Tao Liu
Histone Modification Group, Children’s Cancer Institute Australia,
NSW
1030 – 1045 Patterns of clonal evolution involved in treatment resistance in
Diffuse Large B-cell Lymphoma using tumour and plasma sequencing
Dr Ryan Morin
Simon Fraser University, Canada
Promenade 2 & 3
SESSION 5: CLINICAL SEQUENCING Promenade 1
Chairs: Ms Vivien Vasic and Professor Paul Waring

1115 – 1200 Comprehensive genomic profiling of 5,000+ tumors reveals
new insights into the druggable genomic landscape of solid tumors
Dr Philip Stephens
Foundation Medicine, USA
1200 – 1230 INTEGRATING GENOMICS INTO CLINICAL PRACTICE: A LOCAL AND
INTERNATIONAL PERSPECTIVE
Professor Kathryn North
Murdoch Childrens Research Institute, VIC
1230 – 1300 TRANSFERRING GENOMIC TESTING FROM RESEARCH TO DIAGNOSTICS:
SIMILARITIES AND DIFFERENCES
Professor Hamish Scott
Centre for Cancer Biology, SA

14
2014 AGTA
Conference
TUESDAY
14 October 2014
1300 – 1500 Lunch, Poster session 2 & Trade Exhibition
Meeting Room 3 & 4
CLINICAL SEQUENCING WORKSHOP

The Clinical Sequencing Pipeline: genes to patients
Optional workshop at $50.00 per person. Bookings are essential.
1400 – 1410 Introduction and existing guidelines
Dr Andrew Fellowes
1410 – 1435 NGS TECHNOLOGY: CLINICAL UTILITY AND DESIGN
Dr Desiree duSart
Victorian Clinical Genetics Services, VIC
1435 – 1500 Optimisation and validation
Dr Cliff Meldrum
Hunter New England Health, NSW
1500 – 1525 Standards and Controls
Dr Natalie Thorne
Melbourne Genomics Health Alliance, VIC
1525 – 1550 Afternoon refreshments and Illumina presentation
1550 – 1615 SAMPLE PREPARATION AND SEQUENCING

Professor Graham Taylor
University of Melbourne, VIC
1615 – 1640 BIOINFORMATICS AND COMPUTING IN THE SETTING
OF CLINICAL GENOMICS SETTING
Dr Mark Cowley
Garvan Institute of Medical Research, NSW
1640 – 1705 Curation and reporting of genetic and genomic
data in the modern era (2014!)
Professor Hamish Scott
Centre for Cancer Biology, SA
1705 – 1730 Feedback and discussion
Promenade 1
SESSION 6: TRANSCRIPTOMICS (1500 – 1730)

Chairs: Dr Nicole Cloonan and Dr Nadia Davidson
1500 – 1545 FROM ENCODE BIOCHEMICAL GENOMIC MAPS TO ENHANCER FUNCTION
ASSAYS -- AND BACK: DEFINING THE ACES
Professor Barbara Wold
Caltech, USA
1545 – 1615 IDENTIFYING AND CHARACTERISING TRANSCRIPTS TARGETED BY SUBNUCLEAR-
BODY ASSOCIATED PROTEINS AND LONG NONCODING RNAS
Dr Archa Fox
The Harry Perkins Institute of Medical Research, WA

15
2014 AGTA
Conference
TUESDAY
14 October 2014
1615 – 1645 Unraveling Dendritic Cell-Subset Proudly sponsored by:
Commitment - One Cell At A Time

Dr Andreas Schlitzer
Singapore Immunology Network,
Singapore
1645 – 1700 Recursive splicing enables coupling of promoter choice to splicing
within long vertebrate genes
Dr Christopher Sibley
University College London, England
1700 – 1715 Inferring data-specific micro-RNA function through the joint
ranking of micro-RNAs and pathways from small sample miRNA-Seq
and mRNA-Seq data
Dr Ellis Patrick
University of Sydney, NSW
1715 – 1730 Cross-species Systems Analysis of a Metabolic Disease to Find A Gene
Signature Predictive of Insulin Resistance
Dr Rima Chaudhuri
1900 – 2300 Conference Dinner
ANZ Pavilion, Theatres Building, Arts Centre Melbourne
Optional function at $130.00 per person. Bookings are essential.
The Arts Centre is located on 100 St Kilda Road, Melbourne. An
approximate 15 minute walk from Crown Promenade. Delegates are
asked to make their own way to The Arts Centre for the Conference
Dinner.

16
2014 AGTA
Conference
WEDNESDAY
15 October 2014
0830 – 1530 Registration Desk Open & Welcome Refreshments Promenade

Registraion Desk

Promenade 2 & 3
0855 – 0900 Welcome to Day Three Promenade 1
Promenade 1
SESSION 7: FUNCTIONAL GENOMICS

Chairs: Dr Kaylene Simpson and Dr Amee George
0900 – 0945 EXPLORING IMMUNE SIGNALING SYSTEMS WITH HIGH THROUGHPUT, HIGH
CONTENT SCREENING
Dr Iain Fraser
National Institutes of Health, USA
0945 – 1015 SYNTHETIC LETHAL TARGETING OF E-CADHERIN-DEFICIENT CANCERS
Professor Parry Guilford
University of Otago, New Zealand
1015 – 1030 Replication timing in prostate cancer
Dr Nicola Armstrong
1030 – 1045 Long Non-coding RNA Pathobiology in Colorectal Cancer
Development and Progression
Dr Sheng Liu
Walter and Eliza Hall Institute, VIC
Promenade 2 & 3
Promenade 1
SESSION 8: SINGLE MOLECULE SEQUENCING

Chairs: Dr David Lovell
1115 – 1145 Resolving Genomic Ambiguity: Proudly sponsored by:
Insights into Human Genome

Sequencing and Pathogen
Surveillance using Very Long Reads
Dr Robert Sebra
Icahn School of Medicine, Mount Sinai Institute
for Genomics and Multiscale Biology, USA
1145 – 1300 AGTA Annual General Meeting Promenade 1
1300 – 1345 Lunch & Trade Exhibition Promenade Foyer &

Promenade 2 & 3

17
2014 AGTA
Conference
WEDNESDAY
15 October 2014
SESSION 9: METAGENOMICS Promenade 1
Chairs: Dr Torsten Seemann and Dr Mark Schultz

1345 – 1415 TOWARD ROUTINE CLINICAL SURVEILLANCE OF THE HUMAN MICROBIOTA
Associate Professor Aaron Darling
University of Technology Sydney, NSW
1415 – 1445 THE INFANT AIRWAY MICROBIOME
Dr Kathryn Holt
Bio21 Institute, University of Melbourne, VIC
1445 – 1500 Mapping shared markers across a polynesian
population in the genomic age
Ms Sophia Cameron-Christie
University of Otago, Otago, New Zealand
1500 – 1515 Pore performance? Benchmarking the Oxford Nanopore
Technologies MinION for metagenomics and contig assembly
Dr Ken McGrath
Brisbane Node Manager, The Australian Genome Research Facility,
QLD
1515 – 1530 Awarding of Prizes Promenade 1
2015 Conference Launch

and Conference Close

18
Biology for the post-genomic era
IMPACT
FACTOR
10.5
Editor: Clare Garvey, PhD
• High impact – ranked 6th among research journals in Genetics and Heredity
• Open access for all research
• Fast peer review and rapid publication after acceptance
• High visibility
• Inclusion in PubMed, PubMed Central, Medline and Web of Science
For more information about the journal please visit genomebiology.com
biomedcentral.com
A09435
2014 AGTA
Conference
CONFERENCE
SOCIAL PROGRAM
Welcome Reception Student Functions

Date: Sunday 12 October 2014
Meet-a-Mentor Session
Venue: Crown Promenade, Exhibition Foyer
Time: 1800 – 2000 Date: Monday 13 October 2014
Dress: Business or Smart Casual Venue: Crown Promenade
Time: 1715 – 1815
Enjoy networking with old and new
acquaintances, and familiarising yourself with We invite students and early career researchers
the trade exhibitors, whilst enjoying drinks and to join the conference speakers for scientific
canapés. The Welcome Reception is included speed dating. Like any good speed date, the
in a full registration only. Additional tickets can aim will be to find out the important stuff as
be purchased at $70.00 per person. quickly as possible. Students and early career
researches will have a short time to meet each
mentor to quiz them about how to make it in
Conference Dinner science, how to have a life, and maybe even
make a scientific connection. The bell will ring,
Date: Tuesday 14 October 2014
and you’ll get to do it all over again.
Venue: ANZ Pavilion, Theatres Building,
Arts Centre Melbourne COMBINE Social Event
Time: 1900 – 2300
Dress: Smart Casual Date: Monday 13 October 2014
Venue: PJ O’Brien’s, Southbank
The conference dinner is the social highlight
(10 min walk from Crown Promenade)
of the conference and should not be missed.
Time: 1830 onwards
Come and join us for another chance to
network and meet with colleagues, whilst Students and early career researchers (and
enjoying a great night of food, wine and anyone else keen to join in!) are invited to
dancing. join COMBINE for a fun night of genomics
trivia. Test your knowledge of genomics
technologies, bioinformatics tools, the history
of sequencing and much more! As in real life
genomics, your team will have a diverse range
of backgrounds, so will have to work together
to win the prize! Teams will form on the night,
and points will be issued for both correct and
creative answers.
We will provide finger food and your first few
drinks. Walk with us from the conference venue
at 18:15, just following the Meet-a-Mentor
Session.
Proudly supported by:
20
Handbook
2014 AGTA
Conference
Abstracts and
Biographies
21
Handbook
2014 AGTA
Conference
of Science, Associate Membership of the

SUNDAY European Molecular Biology Organization,
and the Human Genome Organisation’s Chen
12 October 2014
Award for Distinguished Achievement in Human
Genetic and Genomic Research.
Abstract
High throughput analyses have shown that
the vast majority of the human genome
is dynamically transcribed to produce a
ORATION
previously hidden universe of different classes
of small and large, overlapping and interlacing
intronic, intergenic and antisense non-protein-
CHAIR: DR RUBY CY LIN coding RNAs. The transcriptome is far more
complex than the genome, which is best
viewed as a zip file that is unpacked in highly
1700 – 1800 differentiation stage- and cell-specific patterns
during development. These RNAs fulfill a wide
range of regulatory functions, with miRNAs
THE HUMAN GENOME AS THE ZIP FILE and related species being best (although not
EXTRAORDINAIRE well) understood. The functions of the large/
Professor John Mattick AO FAA long noncoding RNAs (lncRNAs) are varied
GARVAN INSTITUTE OF MEDICAL RESEARCH, NSW and include central roles in the formation of
various differentiation-specific subnuclear
Biography
organelles. However, recent evidence suggests
John Mattick is the Director of the Garvan
that the major function of lncRNAs is to guide
Institute of Medical Research. He undertook
chromatin-modifying complexes to their sites of
his undergraduate and graduate training at
action, to specify the architectural trajectories
the University of Sydney and Monash University
of development. Moreover, the coding and
in Melbourne. He subsequently worked at
noncoding transcriptome is far from fully
Baylor College of Medicine in Houston, the
characterized, with RNA CaptureSeq analyses5
CSIRO Division of Molecular Biology in Sydney,
revealing thousands of new exons and isoforms
and the University of Queensland, where he
in human cancer gene loci, and at least 1,500
was the Foundation Director of the Institute
previously unknown lncRNA genes in poorly
for Molecular Bioscience and the Australian
characterized ‘intergenic’ regions that have
Genome Research Facility. He has also
been associated with complex diseases.
spent research periods at the Universities of
Indeed, it is emerging that variations in the
Cambridge, Oxford, Cologne and Strasbourg.
sequence or expression of regulatory RNAs not
John Mattick has published over 250 papers, only underpin phenotypic differences between
which have received over 25,000 citations. His individuals and species, but also play significant
honours include the Honorary Fellowship of the roles in the etiology of diseases. Moreover,
Royal College of Pathologists of Australasia, the emerging transcriptomic, epigenomic
the inaugural Gutenberg Professorship of and nuclear structural data point to an
the University of Strasbourg, the International extraordinary precision of the 4-dimensional
Union of Biochemistry and Molecular Biology organization and expression of the genome
Medal, Fellowship of the Australian Academy that far exceeds current understanding.
22
Handbook
2014 AGTA
Conference
Abstract
Genetic analysis in large structured tree
MONDAY populations using next-generation genomics
13 October 2014 technologies provides a powerful approach
to dissect genetic variation affecting
regulatory networks and biosynthetic pathways
underlying wood development in Eucalyptus
trees. We have performed comprehensive
genome mapping, transcriptome sequencing,
SESSION 1 metabolic profiling and wood cell wall analyses
in an interspecific F2 backcross population of
E. grandis x E. urophylla. Transcriptome profiling
(Illumina RNA-seq) of developing wood of 283
PLANT GENOMICS F2 progeny allowed genome-wide mapping of
expression QTLs (eQTLs) for more than 20,000
CHAIRS: CARSTEN KULHEIM xylem expressed genes. Several trans-eQTL
AND NOEL COGAN hotspots in both backcrosses mark the location
of polymorphisms affecting secondary cell
wall (SCW) biosynthetic genes and provide
0845 – 0930 evidence for segregating components of
the transcriptional network regulating SCW
formation. We further analysed the genetic
POPULATION GENOMICS UNRAVELS GENETIC architecture of transcript level variation for
DIVERSITY AND REGULATION OF GROWTH 353 xylem expressed transcription factors (TFs)
AND DEVELOPMENT IN EUCALYPTUS and show that shared eQTLs and expression
Professor Zander Myburg correlation of TF genes can be used to
UNIVERSITY OF PRETORIA, SOUTH AFRICA reconstruct modules of the SCW transcriptional
network and identify candidates affecting key
Biography
biological processes in xylem development.
Zander Myburg is a professor enetics at the
We also analysed sequence variation in the
University of Pretoria (UP) and holds the Chair
transcriptomes of the F2 progeny and identified
in Forest Genomics and Biotechnology at UP.
266,433 SNPs and 9,682 indels segregating
His research programme in the Forestry and
in the backcross progeny, half of which are
Agricultural Biotechnology Institute (FABI) and
predicted to affect protein sequence and
Genomics Research Institute (GRI) focuses
structure and thereby potentially contribute
on the genomics and molecular genetics
to phenotypic variation. A total of 1,183
of wood development in fast-growing forest
predicted loss-of-function (LoF) mutations are
trees and, in particular, the genetic regulation
polymorphic in the backcross and inactivate
of cellulose biosynthesis in trees. His research
alleles of 1,285 protein-coding genes including
group is pioneering the use of systems genetics
Eucalyptus orthologs of cell wall biosynthetic
approaches to unravel the genetic control
genes. Together these analyses provide the
of wood formation in Eucalyptus trees. He
foundation for gaining a systems genetics
has also been the lead investigator of the
understanding of the regulation of growth and
US Department of Energy (DOE) funded
development in Eucalyptus.
Eucalyptus Genome Project. He has supervised
30 postgraduate (MSc and PhD) students and
is author of 47 peer-reviewed papers and
book chapters in the field of plant molecular
genetics and genomics.
23
Handbook
2014 AGTA
Conference
0930 – 1000 1000 – 1015
PLANT GENOMICS FOR CLIMATE ADAPTION USING SMALL RNA SEQUENCING TO

CONFIRM AND DISCOVER BINDING SITES OF
Associate Professor Justin Borevitz
PENTATRICOPEPTIDE REPEAT PROTEINS
RESEARCH SCHOOL OF BIOLOGY, ANU COLLEGE
OF MEDICINE, ACT Dr Katharine A. Howell
UNIVERSITY OF WESTERN, WA
Biography
Justin obtained his PhD in 2002 from the Biography
University of California at San Diego with Kate Howell is a Research Assistant Professor
Joanne Chory studying Natural Variation in and ARC DECRA Fellow affiliated with the
Arabidopsis light response. His postdoctoral ARC Centre of Excellence in Plant Energy
research was with Joseph Ecker (2002-2004) Biology at The University of Western Australia.
at the Salk Institute focused on genomic She received her PhD from UWA in 2006 and
diversity in Arabidopsis using tiling microarrays. remained at UWA as a Research Associate
Justin then started as assistant and associate until 2007. After receiving an Alexander von
professor in the Dept of Ecology and Evolution Humboldt Foundation Research Fellowship
at the University of Chicago (2004 until 2012). she moved to Germany to work at the Max
His research focused on Genome Wide Planck Institute for Molecular Plant Physiology
Association Studies in Arabidopsis and next in Potsdam. She returned to Perth and
generation genotyping by sequencing in UWA in 2010 and, in 2011, received one of
emerging model organisms. In 2012 Justin the inaugural ARC Discovery Early Career
started at the ANU where current work is Researcher Awards. Her research focuses on
identifing the genetic basis of local adaptation understanding how chloroplasts (the parts of
to seasonal climates using Phenomic and plant cells that perform photosynthesis) make
Landscape Genomic approaches in plant RNA and protein, and communicate with the
model organisms and foundation species. rest of the cell to enable the plant to grow and
Abstract
respond to stress. More recently she has been
Land use and climate change have altered developing methods using next generation
agroecological environments threatening sequencing technologies to examine aspects
food and environmental security. of chloroplast gene expression.
Revolutions in plant genomics, phenomics Abstract
and environmental control are unfolding Pentatricopeptide repeat (PPR) proteins are
with the promise of directly selecting a large family of sequence-specific RNA
regional climate ready seeds for future binding proteins targeted to mitochondria
conditions on marginal land. I will present and chloroplasts that influence the expression
results harnessing natural genetic diversity in of organellar transcripts by altering RNA
models Arabidopsis and Brachypodium along sequence, processing, turnover or translation
with new work in woodland foundation species (Barkan & Small, 2014). While they are found
of switchgrass and Eucalyptus. in all eukaryotes the PPR protein family has
massively expanded in the land plant lineage.
For example, in the plant model, Arabidopsis
thaliana, the family comprises over 450
proteins. However, defining the specific RNA
target(s) of PPRs has been a challenge as
mutations in these proteins often result in
pleiotropic effects.
24
Handbook
2014 AGTA
Conference
Tandem repeats of PPR motifs form helices Abstract
which define the RNA sequence recognised by The exotic fungal pathogen Myrtle rust
the protein. This recognition code has recently (Puccinia psidii) poses a serious biosecurity
been “cracked” (Barkan et al, 2012) allowing threat to ecosystems and forest-based
binding sites of PPRs to be predicted and then industries across the world, including Australia.
tested experimentally. In addition, reanalysis More than 140 Australian native species in
of data generated from small RNA sequencing the family Myrtaceae are susceptible to the
experiments supports the idea that short non- rust disease, including eucalypts (Eucalyptus
coding RNA fragments that accumulate in spp.) and tea trees (Melaleuca spp.). The rust
chloroplasts are likely to be “footprints” of RNA pathogen originated in South and Central
binding proteins (Ruwe & Schmitz-Linneweber, America, and has spread the length of the
2012). These two recent discoveries have the east coast since it was discovered in Australia
potential to rapidly accelerate our ability to in April 2010. The rust affects young leaves,
match PPRs with their RNA targets. flowers and fruits, causing lesions on the host
plant. Finding genetic markers for resistance
We have developed a method to analyse
against Myrtle rust in Myrtaceae may help
putative PPR “footprints” by sequencing small
identify resistant plants and provide a means
RNAs derived from wildtype and mutant plants.
of maintaining production in Myrtaceae-
As a proof-of-concept we have been able to
related rural industries (e.g. oils and culinary
show for 2 PPR proteins, SOT1 and EMB2564,
products such as lemon myrtle) as well as
that putative PPR footprints are lost or reduced
helping nurseries and contributing to restoring
in the mutant plants and that these observed
ecosystems.
changes are consistent with the PPR binding
sites predicted using computational methods. Understanding the molecular basis of
resistance and identifying potential genetic
markers of pathogen resistance in important
in plant breeding programs and we aim
1015 – 1030 to produce molecular markers that will
assist breeding of rust-resistant varieties. We
sequenced the transcriptomes of resistant and
DISCOVERING THE MOLECULAR BASIS OF
susceptible individuals of Melaleuca alternifolia
RESISTANCE IN AUSTRALIAN MYRTACEAE TO
THE EXOTIC FUNGAL PATHOGEN MYRTLE RUST and M. quinquenervia before and five days
(PUCCINIA PSIDII) post-inoculation with Myrtle rust spores. A large
proportion of the assembled transcripts showed
Ms Ji-Fan (Sarah) Hsieh differential expression. We investigated the
THE AUSTRALIAN NATIONAL UNIVERSITY, ACT GO categories of the differentially expressed
Biography transcripts and found many defence-related
Ji-Fan (Sarah) Hsieh is a PhD student in the categories up-regulated in the resistant
Research School of Biology at the Australian plant. One gene family that showed marked
National University. Her research focuses on differences between the resistant and
discovering of the molecular basis of resistance susceptible individuals was the nucleotide-
in Australian Myrtaceae to the exotic plant binding site leucine-rich repeat family (NBS-
pathogen Myrtle rust, and developing LRR), which is responsible for the recognition
molecular markers to help identifying resistant of pathogens and is the start of the signal
plants, providing a means of maintaining cascade that leads to the production of
production in Myrtaceae-related rural industries defence-related proteins and plant secondary
(e.g. oils and culinary products such as metabolites. We are currently developing
lemon myrtle) as well as helping nurseries and molecular markers for resistance based on this
contributing to restoring ecosystems. gene family.
25
Handbook
2014 AGTA
Conference
responsive genes. In order to search for salt-

1030 – 1045 specific genes, we performed transcriptional
profiling in D. salina using HiSEQ2000
technology. Based on 180 million of pair-reads
SALT-SPECIFIC RESPONSIVE GENES ARE
(i.e., 90 base/read), we obtained a de novo
REVEALED BY TRANSCRIPTIONAL PROFILING
OF DUNALIELLA SALINA CELLS IN RESPONSE transcriptome consisting of 24.5 thousand
TO A RECIPROCAL CHANGE OF SALINITY contigs (or ESTs) with the minimal FPKM of 1
that were assembled using Trinity. More than
Professor Jianhua Liu 41% (or 10 thousand) of the 24.5 thousand
INSTITUTE FOR COMPREHENSIVE UTILISATION OF ESTs showed to have at least 1 Best-Hit in the
MARINE BIOLOGICAL RESOURCES, ZHEJIANG, non-redundant protein database (ncbi.nlm.
CHINA
nih.gov) using BLASTX with an cut-off e-value
Biography <1E-03. Off the 10 thousand ESTs, majority ESTs
Dr Jianhua Liu is currently Professor of Ocean had Best-Hits originated from the sequenced
College, Zhejiang University, Hangzhou, green microalgal genomes. The top 2 species
China and Deputy Director of Institute for are Volvox carteri (e.g., 3463 Best-Hits) and
Comprehensive Utilization of Marine Biological Chlamydomonas reinhardtii (e.g., 2846 Best-
Resources of Zhoushan, Zhoushan, China. He Hit). Transcriptional profiling analysis of the
graduated from Fudan University, Shanghai, reciprocal change of salinity (i.e., hyperosmotic
China and received his Ph.D. degree from Free stress: from 0.5M to 2M NaCl and hypoosmotic
University of Brussels, Belgium. Before his job stress: 2M to 0.5M NaCl) indicates that numbers
in Zhejiang University, he was Group Leader of of differentially expressed genes trigged by
the Systems Biology Laboratory in the Genome hyperosmotic stress are 2-fold more than those
Institute of Singapore. His current research triggered by hypoosmotic stress (i.e., 900 versus
efforts focus on analysis of signaling pathways 411 genes at a cut-off of fold-change > 2 and
and transcriptional regulatory networks that p-value < 0.05). We find that 180 genes are
are involved in regulation of cell growth and differentially expressed during both hyper and
response to environmental stress using both hypoosmotic stresses. Correlation coefficient
genetic and genomic approaches in yeast between transcription level changes of these
and microalgae. His laboratory also works on 180 genes upon hyper and hypoosmotic
the marine microbiome structure, function, and stresses is -0.66, suggesting that these genes are
diversity in the Zhoushan Archipelago area. potential candidates of salt-specific responsive
He has published more than 40 peer-reviewed genes. In this meeting, we will report functions
papers. of the potential salt-specific responsive genes
Abstract and its implications in osmotic stress regulation
The green microalga Dunaliella salina can in D. salina.
grow in a wide range of salinities from 0.05M
to 5M of sodium chloride and has become a
useful model for study of cellular adaption to
salinity changes. Previous profiling studies by
others showed that transcriptional changes
occurred in D. salina cells upon salinity increase
or decrease. However, none of the genes
showed induced (or repressed) transcription
upon salinity increase would be repressed
(or induced) when salinity decreases. Genes
displaying the same direction of transcriptional
changes in response to increased or
decreased salinity would not be salt-specific
26
Handbook
2014 AGTA
Conference
Abstract
Cancer involves changes in gene expression
through a variety of mechanisms, including
SESSION 2 translocations and epigenetic alterations.
Recent genome-wide approaches have
revealed a plethora of distal regulatory
elements, some of which have been
EPIGENETICS associated with cancer initiation and
progression. Chromatin interactions, which
CHAIRS: EMMA WHITELAW are two or more distant genomic regions that
AND DAVID MARTINO come together in close spatial proximity in the
3-Dimensional environment of the cell nucleus,
may facilitate cancer through mechanisms
1115 – 1200 such as transcriptional control by connecting
distal regulatory elements with target genes.
3-DIMENSIONAL GENOME ORGANIZATION In this talk, I will present Chromatin Interaction
AND TRANSCRIPTIONAL CONTROL Analysis with Paired-End Tags (ChIA-PET)
sequencing, and present my lab’s recent
Professor Melissa Fullwood advances in refining the ChIA-PET method,
YALE-NUS COLLEGE, SINGAPORE analysis, and application of this method to
Biography better understand transcriptional control in
Dr Melissa J. Fullwood is a Junior Principal cancer cell lines, as well as future goals in terms
Investigator in the Cancer Science Institute with of translation of our results to the clinic.
a joint appointment as an Assistant Professor
at Yale-NUS and a Joint Principal Investigator
position at the Institute of Molecular and Cell 1200 – 1230
Biology, A*STAR, Singapore. Her research
focuses on chromatin interactions and
transcription in cancer. She completed her SHRNA SCREENS FOR NOVEL EPIGENETIC
undergraduate degree in Biological Sciences MODIFIERS AND CHARACTERISATION OF
in 2005 at Stanford University, and her PhD THEIR ROLES IN X INACTIVATION
at the Genome Institute under the NUS
Dr Marnie Blewitt
Graduate School for Integrative Sciences and
WALTER AND ELIZA HALL INSTITUTE, VIC
Engineering in 2009. She was a Lee Kuan Yew
Post-Doctoral Fellow in Duke-NUS. She has 17 Biography
publications, which have been cited over 1500 Marnie Blewitt performed her PhD with Emma
times, and 2 patents. She was one of three Whitelaw developing a mutagenesis screen for
recipients of the inaugural L’Oreal-UNESCO epigenetic modifiers in the mouse, graduating
for Women in Science National Fellowships in in 2005. Marnie worked with Douglas Hilton at
Singapore in 2009, and was the international The Walter and Eliza Hall Institute for her post-
winner of the GE and Science prize in 2010. doctoral studies, where she identified the role
She serves on the editorial board for Scientific ofthe novel protein Smchd1 in X inactivation,
Reports. In 2013, she received the National and studied the function of polycomb group
Research Foundation (NRF) fellowship, which proteins in hematopoietic stem cells. This
comes with a S$3.5 million grant over 5 years. earned her the AAS Gani medal and the
L’Oreal Australia Women in Science fellowship
2009. In 2010, Marnie established her own
group at The Walter and Eliza Hall Institute as
an ARC QEII fellow, to study the molecular
mechanisms of epigenetic control.
27
Handbook
2014 AGTA
Conference
Abstract
Epigenetic marks are frequently reported to 1230 – 1245
correlate with gene silencing, yet there are
few circumstances when the mechanistic role
of specific epigenetic marks in the process CHARACTERISATION OF THE NOVEL
EPIGENETIC MODIFIER RLF
of inactivation have been characterised. X
chromosome inactivation (XCI), the dosage Dr Harald Oey
compensation mechanism in female mammals, LA TROBE UNIVERSITY, VIC
provides a powerful system where the initiation,
Biography
establishment and maintenance of silencing
Harald completed his university studies at
can be studied for hundreds of genes in
Griffith University in 2007. he then worked
parallel. Each phase of silencing can be
for Professor John Mattick at the IMB in
monitored in vitro or in vivo, as female cells
Brisbane for two years before taking on a
transition from two active X chromosomes
postdoctoral research position with Professor
to one active and one inactive. While the
Emma Whitelaw, first at the QIMR in Brisbane
ontogeny of X chromosome inactivation (XCI)
and subsequently at La Trobe University in
is well described, much remains unknown
Melbourne.
about the molecular mechanisms governing
this process. Therefore we are using a bespoke Abstract
shRNA library to screen for epigenetic For many years our lab has been running
regulators of XCI, as in order to understand an ENU mutagenesis screen for modifiers
more about epigenetic silencing, we must first of transgene silencing in the mouse. The
identify all the key players. screen utilizes a GFP transgene expressed
in erythrocytes in a variegated manner
We are using our targeted library in multiple
(expressed in some erythrocytes, silent in
XCI screens: to identify epigenetic regulators
others), and genes that enhance or supress
involved in initiating, establishing or maintaining
variegation are called modifiers of murine
XCI. For high throughput screens, we need a
metastable epialleles (MommeD). Recently,
rapid system to survey XCI. We have produced
we reported the initial characterisation
2 knockin alleles of the X-linked house-keeping
of three novel enhancers of variegation,
gene Hprt: one which produces GFP and the
MommeD8, MommeD28 and MommeD34,
other mCherry under the control of the Hprt
all of which have mutations in a gene called
promoter. We have used these alleles and
Rearranged L-myc fusion (Rlf). Interestingly,
identified the H3K9 methyltransferase Setdb1
while several MommeD alleles cause reduced
as a new regulator of the maintenance of X
DNA methylation at the GFP transgene, and
inactivation. Using a combination of allele-
an increase in GFP expression, Rlf is currently
specific RNA-seq, ChIP-seq and RRBS we have
the only known MommeD causing increased
placed Setdb1 within the epigenetic hierarchy
methylation and a concomitant decrease in
of XCI, which has revealed the broader
GFP expression. To investigate the function
significance of H3K9 methylation in epigenetic
of Rlf further we have carried out a range of
silencing.
experiments to characterise the effect of Rlf
genome-wide. These include whole genome
bisulfite sequencing (~30x coverage) and
mRNA-seq in multiple mutant and wildtype
embryonic tissues, ChIP-seq to identify Rlf DNA
binding sites and co-immunoprecipitation,
followed by mass spectrometry, to identify
Rlf binding partners in-vivo. We find that Rlf
interacts with a histone demethylase in a
28
Handbook
2014 AGTA
Conference
complex known to regulate transcription. methylation to be between the genders, as,

Furthermore, we find that embryos null for even with sex chromosomes excluded from
Rlf have increased DNA methylation at a analysis, samples separate very distinctly
large number of distinct loci throughout the by gender with unsupervised hierarchical
genome and that most sites have chromatin clustering methods. Comparisons of autosomal
marks consistent with roles in transcriptional methylation in 6 male and 6 female mice
regulation. Interestingly, despite widespread showed that females are hypermethylated
effects on the methylome, the effect on the across all chromosomes compared to males.
transcriptome is surprisingly mild. Using a conservative statistical approach we
have identified a list of 221 highly significant
gender-specific differentially methylated
1245 – 1300 regions (DMRs), comprised of 1117 individual

differentially methylated cytosines. Analysis
with gene ontological enrichment tools shows
WIDESPREAD EPIGENOMIC AND GENDER- that these regions are associated with fatty
SPECIFIC DIFFERENCES IN ISOGENIC MICE acid metabolism, iron metabolism, peroxisome
function and muscle differentiation, among
Ms Helen McCormick
other processes. Furthermore, genes associated
VICTOR CHANG CARDIAC RESEARCH INSTITUTE, NSW with these DMRs are primarily non-tissue
Biography specific, suggesting that these differences may
Helen McCormick is a PhD student from the be established in very early development.
epigenetics laboratory at the Victor Chang
Cardiac Research Institute in Sydney. Her
research focuses on the nature and extent
of transgenerational epigenetic inheritance
in mammals, as well as gender-specific
methylation. Prior to this she completed her
MSc at the University of Waikato in Hamilton,
New Zealand on molecular ecology of wild
mice.
Abstract
Epigenetic variation between individuals
can provide a mechanism for phenotypic
variation and differences in disease risk. While
a proportion of epigenetic variation can
be ascribed to genetic factors, epigenetic
variation can persist even in the absence of
genetic differences. To investigate epigenetic
variation without genetic heterogeneity,
we have performed genome-wide DNA
methylation analysis on genetically identical
mice. We have compared sibling groups
from different families and find that, while
all mice display inter-individual epigenetic
variation, closely related animals are more
similar to each other than to those that are
more distantly related. However, we find
the largest differences in genome-wide
29
Handbook
2014 AGTA
Conference
1600 – 1630
SESSION 3
The STATegra project: new statistical
tools for analysis and integration of
diverse omics data.
BIOINFORMATICS Dr Ana Conesa

PRINCE FELIPE RESEARCH CENTER, VALENCIA,
CHAIRS: GORDON SMYTH SPAIN
AND DAVID GOODE Biography
Dr Conesa is head of the Genomics of Gene
Expression Lab at the Prince Felipe Research
1515 – 1600 Center is Valencia, Spain. She has a degree
in Agricultural Engineering and a PhD in
molecular microbiology from the University of
WHAT’S AHEAD FOR BIOLOGY?
Leiden of the Netherlands. From the year 2000
THE DATA-INTENSIVE FUTURE
she has developed her scientific career further
Associate Professor Titus Brown as bioinformatician where she has specialized
MICHIGAN STATE UNIVERSITY, USA in transcriptomics.
Biography Her research interests split between the

C. Titus Brown is an assistant professor in functional annotation of sequence data, and
the Department of Computer Science the development of statistical methods for
and Engineering and the Department of the analysis of differential gene expression
Microbiology and Molecular Genetics. He and data integration. She has published over
earned his PhD (‘06) in developmental 60 scientific papers and developed popular
molecular biology from the California software tools like Blast2GO and Qualimap,
Institute of Technology. Brown is director web-based applications (Paintomics, SEA)
of the laboratory for Genomics, Evolution, and R packages such as maSigPro and
and Development (GED) at Michigan State NOISeq. Currently her research focuses on
University. He is a member of the Python computational approaches for the functional
Software Foundation and an active contributor annotation on long non-coding RNAs and the
to the open source software community. His integrative modeling of functional NGS data
research interests include computational for biomedical applications.
biology, bioinformatics, open source software At present she is scientific coordinator of the
development, and software engineering. EU projects STATegra in omics data integration
Abstract and the Marie Curie DEANN for the creation
The rise of -omics over the last 20 years, and the of the NGS network between Latin-American
increasing need for efficient and integrative and European labs. She is also Management
approaches to data analysis, portends some Committee member of the EU COST Action
interesting and challenging changes in biology SeqAhead.
research over the next decade.
For one example, what happens when high-
density multi-omics data is available for low
cost for every experiment? I will discuss both
my own work on efficient sequence analysis
approaches as well as a larger context of some
larger-scale data investigations.
30
Handbook
2014 AGTA
Conference
Abstract
Next generation sequencing has speed
1630 – 1645
up genome analysis and brought omics
research closer to many organisms and CLASSIFICATION OF GENOMIC FUSIONS INTO
biological scenarios. Today an increasing STRUCTURAL VARIATION EVENTS
number of research projects propose the
combined use of different omics platforms Dr Jan Schroeder
to investigate diverse aspects of genome WALTER AND ELIZA HALL INSTITUTE, VIC
functioning. These proposals ideally seek to Abstract
provide complementary sources of molecular A precise understanding of the structural
information that eventually can be put variations (SVs) between samples of DNA is
together to obtain systems biology models important in the study of population diversity
of biological processes. Hence, it is not rare and disease causing mutations. There are
anymore to find experimental designs involving numerous methods that are designed to
the collection of genome, transcriptome, detect SVs in DNA sequencing data. Most of
epigenome and even metabolome data these methods do not provide a higher-level
on a particular system. However, standard interpretation of variants, instead identifying
methodologies for the integration of diverse a large set of coordinate pairs that are
omics data types are not yet ready and connected via candidate fusions – so-called
researchers frequently face post-experiment break points. While some SV callers include a
question on how to combine data of different description of underlying genomic events, they
nature, variability, and significance into an are currently limited to a set of simple variant
analysis routine that sheds more light than types (such as deletions or inversions).
the analysis of individual datasets separately.
Here we present an algorithm to process
The STATegra project has been conceived
the output of SV calling algorithms and infer
to address these problems and provide the
the rearrangement events that explain the
genomics community with user-friendly tools for
observed fusions. Our method creates a graph
the integration of different omics data types.
data structure from the fusion information
STATegra targets several sequencing based
and looks for patterns that are characteristic
functional genomics methods, proteomics
of more complex rearrangement types (such
and metabolomics. In this presentation I will
as balanced translocations). Our algorithm
describe current results of the project including
summarises fusion events that are seemingly
the STATegraEMS, an experiment management
independent in the original output into single,
system for storage and annotation of complex
more complex events (for example, an intra-
omics experiments, experimental design issues,
chromosomal translocation involves three
novel integrative visualization tools, statistical
fusions) thus reducing the output to a more
approaches to integrate RNA-seq data with
intelligible set. A better-categorised output
other omics technologies, and data mining
allows for better filtering of specific events
strategies to leverage public domain datasets
that are most relevant to the experimental
in the integrative effort. I will also present the
circumstances.
STATegRa, a new Bioconductor R package for
integrative omics data analysis. We demonstrate that our method can improve
the output of various SV algorithms (such as
CREST, BreakDancer, Delly, and Socrates),
identifying redundant events, and reducing
false positives. Augmenting SV calls in this
manner simplifies exploratory SV analysis of
samples in various experimental settings,
such as whole genome data, exome data, in
somatic or germline tissue.
31
Handbook
2014 AGTA
Conference
transcriptome assembly. Validation on RNA-Seq

1645 – 1700 from breast-cancer cell lines and simulation,
show that JAFFA outperforms competing
methods.
IDENTIFYING FUSION GENES AND OTHER
TRANSCRIPTIONAL VARIATION IN CANCER In addition to fusions, tumour transcriptomes
THROUGH DE NOVO TRANSCRIPTOME harbour information on other variants of
ASSEMBLY potential oncogenic significance such as SNPs,
indels and mis-splicing. Based on the idea
Dr Nadia Davidson
of using de novo transcriptome assemblies
MURDOCH CHILDRENS RESEARCH INSTITUTE, VIC
we also present new work to identify these
Biography transcribed variants in a cancer sample. Our
Dr Nadia Davidson is a bioinformatician who results are independent of either a reference
has been working within the Oshlack group at genome or transcriptome, because each
the Murdoch Childrens Research Institute since cancer transcriptome within a cohort is
2011. She was originally trained in Physics and compared to all other transcriptomes. We
Software Engineering and completed her PhD show examples of specific genes such as tp53
in 2011 from the University of Melbourne on the that have unique and previously unobserved
topic of Experimental Particle Physics at the transcripts in some tumours.
Large Hadron Collider. Her research interests
include methodology development and
analysis of next generation RNA sequencing
data. She has been involved in a diverse set of
1700 – 1815
projects that include studying sex development
in birds, to identifying genomic rearrangements DEVELOPING AN NGS PIPELINE FOR
in cancer. A common theme to her work is de DIAGNOSTICS
novo transcriptome assembly.
Mr Liam McIntyre
Abstract
SA PATHOLOGY, SA
Cancer is a disease in which lesions of the
genome drive tumour development. One Biography
class of genomic lesion is a fusion gene - Liam graduated with a Bachelor of Science
whereby two genes are joined as a result (Biochem/Biomed) at the University of
of a translocation, producing a new and Queensland in 2007 and then completed his
potentially oncogenic gene. Many fusions Masters in Bioinformatics at the University of
are recurrent in patient tumour populations Queensland in 2014. Currently employed by SA
and their identification has both research pathology as a genetic informatician.
and clinical benefits. RNA sequencing (RNA- Abstract
Seq) is a powerful method for profiling cancer Next-generation sequencing (NGS) has
transcriptomes to identify fusions. However, reached a price point where it is viable for
many methods for identifying fusions have low clinical diagnostic application, however,
precision, require tuning of parameters or are the development of bioinformatics pipelines
not suitable for all RNA-Seq read lengths and remains an active area of research. We have
layouts. developed a read alignment and variant
Here, we present JAFFA (https://code. calling pipeline based on BWA-MEM and
google.com/p/jaffa-project/), a method that the GATK best practices and have adapted
overcomes these shortfalls. JAFFA is based on it to suit amplicon and enrichment based
the idea of comparing the sequenced cancer protocols for sequencing on the MiSeq. We
transcriptome to a reference transcriptome, have compared its performance against the
such as GENCODE. For short reads, the vendor pipeline, MiSeq Reporter, by assessing
cancer transcriptome is built through de novo recall rate and specificity using retrospective
32
Handbook
2014 AGTA
Conference
and prospective sample analysis. We found

BWA-MEM achieves higher percentage of
mapped reads and coupled with GATK’s TUESDAY
haplotype caller achieves higher recall rates 14 October 2014
and accuracy, especially for indels. Use of
the GVCF mode in GATK reduces the total
compute time over earlier versions of the
software. Further, we have developed an
in-house QC pipeline to assess target region SESSION 4
coverage. We use this pipeline to characterise
coverage for a given gene panel and to define
regions commonly failing coverage, which are
often associated with high GC content; we
CANCER GENOMICS
further perform detailed coverage analysis
on each sample. Clinically, it is important to
CHAIRS: MARK WALTHAM
report not only on the confidence of a given
AND IAN MAJEWSKI
variant call, but also on the confidence of
every negative call in every gene of interest.
Remaining challenges for our pipeline are
0830 – 0915
calling variants in genes with pseudo gene
copies, especially where there is recombination
between the functional and pseudo gene INTERROGATING THE ARCHITECTURE OF
copies and continuing to improve our ability to CANCER GENOMES
call indels, especially those larger than 10bp.
Dr Peter Campbell
WELLCOME TRUST SANGER INSTITUTE, CAMBRIDGE,
UK
Biography
Dr Peter Campbell is Head of Cancer Genetics
and Genomics at the Wellcome Trust Sanger
Institute, having started a Wellcome Trust Senior
Clinical Fellowship in 2010. He completed
specialist training in Haematology in 2002.
Following this, he completed a PhD at the
University of Cambridge in the molecular
pathogenesis and clinical management of
myeloproliferative disorders. Since 2007, Dr
Campbell has been employed at the Cancer
Genome Project, Wellcome Trust Sanger
Institute. In 2010, he was awarded the CR-UK
Future Leaders in Cancer Research Prize.
His major interest is cancer genomics, and his
recent research has been concentrated on the
implementation of next generation sequencing
technologies for the detection of somatically
acquired genetic variants in tumour samples.
One major aspiration is to develop translational
applications of high-throughput genomic
screening for the care of patients with cancer.
Further details are available at: http://www.
sanger.ac.uk/research/faculty/pcampbell/
33
Handbook
2014 AGTA
Conference
Abstract largest population-based cohort studies of

Cancer is driven by mutation. Using massively ovarian cancer in the world, involving almost
parallel sequencing technology, we can 3000 women. Professor Bowtell’s lab focuses
now sequence the entire genome of on the genomic analysis of ovarian cancer
cancer samples, allowing the generation and the genetic basis of treatment resistance.
of comprehensive catalogues of somatic His work is part of the International Cancer
mutations of all classes. Bespoke algorithms Genome Consortium, an ambitious program
have been developed to identify somatically to map the genomes of the 50 most common
acquired point mutations, copy number cancer types.
changes and genomic rearrangements,
Abstract
which require extensive validation by
It has become increasingly apparent that
confirmatory testing. The findings from our first
invasive epithelial ovarian cancer is a series of
handful of genomes illustrate the potential
entirely distinct diseases that individually bear
for next-generation sequencing to provide
a closer relationship to other solid cancers
unprecedented insights into mutational
than each other, and that many arise from
processes, cellular repair pathways and gene
non-ovarian sites. Therefore, a comprehensive
networks associated with cancer development.
reclassification of ovarian cancer is needed,
I will also review possible applications of these
based on underlying molecular features, and
technologies in a diagnostic and clinical
the development of treatments directed
setting, and the potential routes for translation.
towards specific subtypes.
Our work draws on patient samples accrued
from the Australian Ovarian Cancer Study
0915 – 0945 (AOCS), involving ~3000 women presenting
for surgery for ovarian cancer or who have
recurrent disease, and more recently from
THE GENOMIC LANDSCAPE OF PRIMARY AND
a rapid autopsy study (CASCADE). We have
ACQUIRED RESISTANT HIGH GRADE SEROUS
particularly focused on high grade serous
OVARIAN CANCER.
ovarian cancers (HGSC), which account for
Professor David Bowtell ~60% of ovarian cancer deaths, and seek to
PETER MACCALLUM CANCER CENTRE, VIC understand factors that impact on the clinical
Biography
outcome of HGSC patients.
After 10 years as Director of Research at The presentation will discuss molecular
Peter MacCallum Cancer Centre, Melbourne, subtypes of HGSC and their association with
Professor Bowtell returned to full time research in drug response and survival. I will discuss our
early 2010. He is Head of the Cancer Genomics recent unpublished studies involving the whole
and Genetics Program and a Group Leader at genome sequencing of 92 patients, including
Peter Mac. paired samples collected pre-treatment and
Professor Bowtell trained as a molecular post-resistance, providing the largest whole
biologist and has an extensive background genome study of ovarian cancer to date. I
human cancer genomics and signal will describe novel mutations associated with
transduction. He is Principal Investigator for the acquired drug resistance.
Australian Ovarian Cancer Study, one of the
34
Handbook
2014 AGTA
Conference
tumor specific DNA present in the circulation,

0945 – 1015 however, the application of next generation
sequencing technologies is now providing
new opportunities to develop ctDNA as a
CIRCULATING TUMOUR DNA AS A LIQUID
non-invasive ‘liquid biopsy’ alternative to
BIOPSY IN CANCER
tissue biopsies for use in cancer diagnostics
Dr Sarah-Jane Dawson and management. Our recent research has
PETER MACCALLUM CANCER CENTRE, VIC revealed the high sensitivity of ctDNA as a
Biography
molecular biomarker for disease monitoring in
Dr Sarah-Jane Dawson is a consultant breast cancer, and demonstrated the novel
oncologist and a clinician-scientist who application of serial plasma DNA analysis to
obtained her medical degree from the study mechanisms of resistance to cancer
University of Melbourne, Australia. She recently therapies. I will present an overview of these
completed postdoctoral studies under the recent and exciting developments as well as
direction of Prof. Caldas and Dr Rosenfeld an update on current ongoing research.
at the CRUK Cambridge Institute, University
of Cambridge, UK and is now the head of
the newly formed Molecular Biomarkers and 1015 – 1030
Translational Genomics Laboratory at the Peter
MacCallum Cancer Centre in Melbourne. Her
current research interests lie in understanding A NOVEL LONG NONCODING RNA,
the genomic evolution of cancer and using LNCUSMYCN, PROMOTES TUMOURIGENESIS
this information to develop noninvasive BY BINDING TO THE RNA-BINDING PROTEIN
molecular biomarkers for clinical application. NONO AND UP-REGULATING MYCN
Her recent research has centered on the ONCOGENE EXPRESSION
application of next generation sequencing to Dr Tao Liu
study circulating cell free tumor DNA (ctDNA). CHILDREN’S CANCER INSTITUTE AUSTRALIA, NSW
This work has resulted in two landmark studies
Biography
demonstrating the high sensitivity of ctDNA as
Dr Tao Liu obtained his PhD from The University
a molecular biomarker for disease monitoring
of New South Wales in 2000. He then worked
in breast cancer, and showing the novel
as a post-doc with Professor Samuel Breit
application of serial exome sequencing of
at St. Vincent’s Centre for Applied Medical
plasma to study mechanisms of resistance to
Research Sydney and Professor Glenn Marshall
cancer therapies. Together, these studies have
at Children’s Cancer Institute Australia. He
established a paradigm for the use of ctDNA
is now Head of Histone Modification Group
as a ‘liquid biopsy’ for molecular disease
at Children’s Cancer Institute Australia, and
monitoring in solid malignancies, and as an
an Australian Research Council Future Fellow
alternative to tissue biopsies for the study of
at The University of New South Wales. In
tumor evolution during the course of disease
the last 5 years, Dr Liu has been working on
and treatment.
the roles of long noncoding RNAs, histone
Abstract
methyltransferases and histone deacetylases
Cell-free circulating DNA containing tumor- in modulating gene transcription, N-Myc-
specific sequences can be identified in the regulated malignant transformation,
plasma of cancer patients. Serial analysis neuroblastoma initiation and progression in
of ctDNA can allow the evolving genomic vitro and in vivo, as well as the anticancer
landscape of a cancer to be assessed, with efficacy of antisense oligonucleotides
many potential clinical applications. Analysis targeting long noncoding RNAs, histone
of ctDNA is challenging and requires highly methyltransferase inhibitors and histone
sensitive techniques due to the small fraction of deacetylase inhibitors in vitro and in vivo.
35
Handbook
2014 AGTA
Conference
Abstract
Patients with neuroblastoma, due to the 1030 – 1045
amplification of a 130 kb genomic DNA region
containing the MYCN oncogene, have a
PATTERNS OF CLONAL EVOLUTION INVOLVED
survival rate of less than 10%. While MYCN
IN TREATMENT RESISTANCE IN DIFFUSE LARGE
has been extensively studied in the last three
B-CELL LYMPHOMA USING TUMOUR AND
decades, it is unknown whether other genetic
PLASMA SEQUENCING
elements within the 130 kb amplicon contribute
to tumorigenesis. The novel long noncoding Dr Ryan Morin
RNA up-stream of MYCN (lncUSMycN) in SIMON FRASER UNIVERSITY, CANADA
the 130 kb amplicon has been recently Biography
manually annotated by the Human and Dr Morin is an Assistant Professor, co-appointed
Vertebrate Analysis and Annotation (HAVANA) in the department of Molecular Biology and
bioinformatics team, but has not been Biochemistry at Simon Fraser University and
studied. Using 5’- and 3’- rapid amplification the BC Cancer Agency's Genome Sciences
of cDNA ends PCR, we experimentally Centre. He earned both M.Sc. and Ph.D.
identified 140bp extra nucleotides at the degrees from the University of British Columbia
5’- end of lncUSMycN. Analysis of 341 human (Canada) in the laboratory of Marco Marra.
neuroblastoma samples showed that the Dr Morin is best known for his early work using
lncUSMycN gene was co-amplified with the massively parallel sequencing to identify
MYCN oncogene. RNA-binding protein pull- driver mutations in leukemias and lymphomas
down assays and mass spectrometry analysis including histone modifiers such as EZH2 and
identified the RNA-binding protein NonO MLL2. Some of his recent recent work has been
as a lncUSMycN RNA binding protein, and published in world-class journals such as the
RNA-immunoprecipitation assays showed New England Journal of Medicine, Nature
NonO bound to both lncUSMycN and Genetics, Cancer Cell and Nature. He has
MYCN RNAs. While knocking-down NonO contributed to 42 peer-reviewed articles and
and lncUSMycN with small interfering RNAs 5 reviews/book chapters on various aspects
reduced MYCN RNA expression, forced over- of genomics and DNA sequence analysis and
expression of lncUSMycN resulted in MYCN his published work has been collectively cited
RNA up-regulation by binding to NonO. In 476 over 5500 times.
human neuroblastoma samples, high levels
Abstract
of lncUSMycN and NonO RNA expression
DLBCL is a common aggressive non-Hodgkin
correlated with high levels of MYCN RNA
lymphoma (NHL) and although sequencing
expression and independently predicted
diagnostic specimens has identified many
poor patient prognoses. Moreover, treatment
driver genes, it is unclear whether additional
with antisense oligonucleotides targeting
genes are involved in relapse or resistance
lncUSMycN in neuroblastoma-bearing mice
to standard treatments (R-CHOP). We
significantly reduced MYCN expression
performed a meta-analysis of all published
and blocked tumor progression. Our data
data interrogating the genomes of DLBCLs.
therefore demonstrate the important role of
To determine the extent to which clonal
the novel long noncoding RNA lncUSMycN in
evolution contributes to treatment resistance,
regulating MYCN oncogene expression and
we characterized DLBCLs following relapse
neuroblastoma oncogenesis, and provide
using exome and genome sequencing. We
the first evidence that amplification of long
have also implemented approaches to detect
noncoding RNA genes can contribute to
mutant cell-free DNA in the plasma known
tumorigenesis.
as circulating tumour DNA to monitor tumour
evolution. We analysed ctDNA using deep
amplicon sequencing and targeted capture
to non-invasively detect evolution of DLBCL
36
Handbook
2014 AGTA
Conference
tumours. By comparing diagnostic tumours and

ctDNA collected at relapse, we have observed
examples of branched clonal evolution in
which each clone appears to derive from a SESSION 5
common progenitor clone with a restricted
set of initiating mutations and a unique set
of additional mutations. By comparing the
prevalence of mutations in individual genes CLINICAL SEQUENCING
in relapse biopsies to those in prior studies of
diagnostic tumours, we have identified multiple CHAIRS: VIVIEN VASIC
genes that are enriched for mutations following AND PAUL WARING
relapse. Retrospective analysis of earlier
biopsies confirms the sub-clonal representation
of these mutations, suggesting that treatment 1115 – 1200
has selected for populations bearing specific
mutations. This study demonstrates branched
clonal evolution in DLBCL including acquisition COMPREHENSIVE GENOMIC PROFILING OF
of additional driver mutations during treatment 5,000+ TUMORS REVEALS NEW INSIGHTS INTO
that are not clonally abundant at diagnosis. THE DRUGGABLE GENOMIC LANDSCAPE OF
We are actively investigating the modes SOLID TUMORS
by which these genes and mutations may Dr Philip Stephens
contribute to tumour aggressiveness and FOUNDATION MEDICINE, USA
resistance to R-CHOP.
Biography
Dr Stephens joined Foundation Medicine in
March 2011, bringing more than a decade
of experience in cancer genomics to the
company. Dr Stephens is a world-renowned
expert in next-generation sequencing and
cancer genome analysis and has authored
numerous publications in Nature, Nature
Genetics, Nature Medicine, Cell and other
high-profile journals.
Prior to joining Foundation Medicine, Dr
Stephens held various senior research positions
during his 11-year tenure with the Cancer
Genome Project at the Wellcome Trust Sanger
Institute under the direction of Professor
Michael Stratton. During this time, Dr Stephens
was a member of the team that sequenced
the first two comprehensive melanoma and
lung cancer genomes, and was co-lead author
in the discovery of BRAF in melanoma and
ERBB2 in lung cancer.
Dr Stephens received a PhD from Oxford
University.
37
Handbook
2014 AGTA
Conference
Abstract
Background: Characterization of genomic 1200 – 1230
changes driving an individual’s disease is
essential to inform rational use of targeted
INTEGRATING GENOMICS INTO CLINICAL
therapies and treatment planning for many
PRACTICE: A LOCAL AND INTERNATIONAL
patients with cancer. We have developed
PERSPECTIVE
a comprehensive, pan-solid tumor next
generation sequencing (NGS)-based Professor Kathryn North
diagnostic assay, optimized for routine FFPE MURDOCH CHILDRENS RESEARCH INSTITUTE, VIC
specimens including fine needle biopsies/
Biography
aspirates and applied it to 5,800 patients’
Kathryn is the Director of the Murdoch Childrens
tumors in a CLIA-certified, CAP-accredited lab.
Research Institute, Director of the Victoria Clinical
Methods: Hybridization capture of 3,769 Genetics Service and the David Danks Professor
exons from 236 cancer-related genes and 47 of Child Health Research at the University of
introns of 19 genes commonly rearranged was Melbourne.
applied to ≥20ng of DNA extracted from 5,800
Kathryn trained as a paediatrician, neurologist
consecutive FFPE specimens and sequenced to
and clinical geneticist in Sydney and Boston.
a unique median depth >1,000X.
Her laboratory research interests focus on gene
Results: 95% (5,508) of specimens were discovery and disease mechanism in inherited
successfully profiled, yielding 2,711 unique muscle disorders, as well as genes which
actionable alterations across 82.1% (4,520) influence normal skeletal muscle function and
of cases. 68.0% (3,744) of tumors harbored elite athletic performance. Her clinical research
≥1 actionable alteration not assayed by focuses on clinical trials for muscular dystrophy
available tumor-type specific tests or hotspot and neurofibromatosis type 1.
panels. Several novel, recurrent fusions
Kathryn is Chair of the NHMRC Research
were identified including NTRK1, FGFR2 and
Committee, and in 2012 was awarded the
BRAF fusions in pan-negative lung cancer,
Ramaciotti Medal for Excellence in Biomedical
cholangiocarcinoma, and melanoma,
Research and the Member of the Order of
respectively. Pan-negative melanoma was also
Australia (AM) for service to medicine in the
significantly enriched for NF1 loss. Constitutively
field of neuromuscular and neurogenetics
activating mutations in ESR1 were identified
research. Kathryn is one of the leaders of
as a novel resistance mechanism to hormonal
Melbourne Genomics Health Alliance, a
therapy in up to 20% of patients and ERBB2
collaborative approach to the integration of
activating mutations were enriched in 40% and
genomic information into everyday healthcare.
27.3% of micropapillary urothelial carcinoma
Melbourne Genomics is a partnership of the
and CDH1-mutated lobular breast cancer,
Royal Melbourne Hospital, Royal Children’s
respectively. Responses to targeted therapies
Hospital, Murdoch Childrens Research Institute,
against actionable alterations identified in
Walter and Eliza Hall Institute, CSIRO, Australian
patient tumors have been reported by treating
Genome Research Facility and the University
physicians.
of Melbourne. In 2013, Kathryn was invited to
Conclusions: Comprehensive genomic profiling participate in the Global Alliance for Genomics
identified actionable alterations in the majority and Health, an international consortium of more
of tumors profiled and identified potential than 130 institutions across 40 countries exploring
additional treatment options for 68% of patients the sharing of genomic and clinical data –
targeting alterations in genes not currently Kathryn is Vice Chair of the Steering Committee
assayed by available hotspot genotyping and Co-Chair of the Clinical Working Group. The
panels. Global Alliance will play a pivotal role in how
genomic data can be linked and shared on a
global level, and applied to the development of
targeted therapeutics.
38
Handbook
2014 AGTA
Conference
Abstract
disease and patient advocacy organizations.
Genomic technologies are proving
The Global Alliance aims to accelerate the
transformative, but ensuring full clinical
world-wide effort to responsibly aggregate,
and research benefit from their application
analyse and share large amounts of genomic
in clinical settings requires planning and
and clinical information to advance the
collaboration. The Melbourne Genomics
understanding, diagnosis, and treatment for
Health Alliance (MGHA) is a collaboration of
cancer, inherited diseases, infectious diseases,
10 research and healthcare organisations, with
and drug responses.
the goal of integrating genomic information
into everyday healthcare. MGHA are
conducting a clinically-led, prospective project
evaluating the feasibility of whole exome 1230 – 1300
sequencing (WES) as a single, first tier assay for
germline and somatic conditions. The Alliance
members have developed common ethics TRANSFERRING GENOMIC TESTING FROM
RESEARCH TO DIAGNOSTICS: SIMILARITIES
and consent, a common clinical bioinformatics
AND DIFFERENCES
pipeline and share genomic data via a
common clinical genomics data repository. Professor Hamish Scott
Decision-making is guided by structured CENTRE FOR CANCER BIOLOGY, SA
processes for clinician input and a Community
Biography
Advisory Group. Researchers are also an
Professor Hamish Scott did his PhD (1992) and
integral part of advisory and working groups.
first post-doc at the Women’s and Children’s
Mimicking usual clinical practice, patients Hospital and the University of Adelaide.
with one of five diverse germline or somatic During these 7 years he led the discovery of
conditions are being offered whole exome genes for 3 rare human diseases. After 11
sequencing in parallel to routine investigations. more years, with the persistence of Professor
Sequence data is generated by multiple John Hopwood and others in academia and
diagnostic laboratories, analysis is then industry, this resulted in either FDA approved
targeted to genes known to be related to the therapy (2003) or clinical trials of novel
clinical condition using a common analytic therapies for these diseases.
pipeline. Research results are returned to
In 1995, Hamish moved to the University of
clinicians and data is available to researchers.
Geneva Medical School in Switzerland. His
Exome data is linked to clinical data and a
focus was, and remains, the application
consolidated electronic view provided to
of genetic and genomic technologies to
clinicians and researchers. The pilot phase is
understand diseases processes to improve
being evaluated to determine the barriers,
diagnoses and treatment. He led international
feasibility, health economics and diagnostic
collaborations in identification of human genes
value of genomic sequencing. This systematic
causing Down syndrome and rare forms of
approach is designed to foster incremental
genetic deafness and autoimmunity (e.g.
change and future adoption, as well as
arthritis and multiple sclerosis). This continues to
ensure future implementation delivers a
have profound effects on our understanding
viable and sustainable system across multiple
of basic biology of Down syndrome, hearing
organisations.
and the immune system and lead to new
The Melbourne Genomics Health Alliance therapeutic strategies in these and related
is also a member of the Global Alliance diseases. This was also the start of his interest
for Genomics and Health (GA4GH), an in cancer and leukemia as children with
organisation of over 200 of the world’s leading Down syndrome have a low incidence of solid
biomedical research institutions, healthcare tumours and a high incidence of leukemia.
providers, information technology and life This is also when he started to work on familial
science companies, funders of research, and predisposition to leukemia.
39
Handbook
2014 AGTA
Conference
Hamish relocated to the Walter and Eliza Abstract

Hall Institute of Medical Research (WEHI) in The gap between research genetic testing
Melbourne in 2000 as the Inaugural Nossal and diagnostic genetic testing is arguably
Leadership Fellow. He was appointed as a closing at a rapid rate due to next generation
National Health and Medical Research Council sequencing (NGS). This talk will highlight some
(NHMRC) senior research fellow in 2001. His of the similarities and differences between
laboratory identified 2 additional deafness the two from the perspective of a consumer,
genes and described the role of a gene in referring and treating clinicians, research and
reprogramming the DNA of an “adult” cell diagnostic scientists. We will discuss examples
from “normal” to become a gamete (sperm or of the current implementation of various NGS
oocyte). He also described a cause of familial panels for both germline and somatic testing at
predisposition to leukemia. SA Pathology.
Since returning to Adelaide in January 2008 Many NGS test have technical (laboratory
he has been Deputy and then Head of the and bioinformatic) limitations as to what type
Department of Molecular Pathology at SA of mutations they will actually detect, that
Pathology, He is an inaugural member of the also affect the diagnostic yield, sensitivity and
Centre for Cancer Biology an Affiliate Professor specificity of the tests. For germline testing we
in both the Schools of Medicine and Molecular have trialled panels, Whole exome sequencing
and Biomedical Science at the University of (WES) and Whole Genome Sequencing (WGS).
Adelaide and an Adjunct Professor in the Across 152 germline samples in our validation
School of Pharmacy and Medical Sciences in study of NGS panels we determined an overall
the Division of Health Sciences of the University sensitivity for SNV and indel detection (380 SNPs
of South Australia. He is an NHMRC principal and pathogenic variants) of 0.9, but identified
research fellow and a Founding Fellow of gaps in coverage, often associated with GC
the Faculty of Science (FFSc) of the Royal rich exons. Genes undergoing recombination
College of Pathologists of Australasia (RCPA). with pseudo-gene copies eluded our current
He recently led the identification of mutations analysis by NGS. After manual review in IGV,
in a gene in rare families and patients that there were no false positives. NGS testing
predisposed them to acute myeloid leukemia of somatic tumor samples further needs to
(AML), infectious diseases and lymphoedema. consider DNA quality (e.g. from FFPE samples)
He has helped develop and introduce and amount of DNA available, especially for
new technologies and tests for improved biopsies from lung samples. Different panels
treatment (personalized medicine) and is a with different clinical inputs have dramatically
Joint Director of the ACRF Cancer Genome different diagnostic yields and levels of
Facility established at SA Pathology. In these detection of variants of unknown significance
roles he has been central to introducing both which may dramatically alter follow-up testing
somatic and germline genotyping using next required or the reporting of a negative result.
generation sequencing to SA Pathology at a We will discuss two cases where standard
panel, exome and whole genome level. genetic testing, and exome analyses in one
case, had failed to reach a diagnosis but WGS
has at least provided a partial clinical solution
resulting in altered patient treatments. For
somatic testing, we will give research examples
from CML, AML and MDS, showing that
sensitivities of detection down to 1/104 cells is
clinically significant while a widely accepted
diagnostic standard is just 1/20. Interpretation
of both germline and somatic results remains
key, particularly for diagnostic reporting,
clinical interpretation and action.
40
Handbook
2014 AGTA
Conference
Australasia. He was the founding chairman of

the Molecular Genetics Society of Australasia
in 2005, the molecular genetics special interest
WORKSHOP group of the Human Genetics Society of
Australasia (HGSA). He became a Fellow of the
HGSA in the discipline of Molecular Genetics
in 2008 and a founding fellow of the Faculty of
CLINICAL SEQUENCING workshop Science of the Royal College of Pathologists of
Australasia (RCPA) in 2011.
The Clinical Sequencing Pipeline:
genes to patients
1410 – 1435
Optional workshop at $50.00 per person.
Bookings are essential.
NGS TECHNOLOGY:
1400 – 1410 CLINICAL UTILITY AND DESIGN
Dr Desirée du Sart
VICTORIAN CLINICAL GENETICS SERVICES, VIC
INTRODUCTION AND EXISTING GUIDELINES
Biography
Dr Andrew Fellowes
Desirée du Sart: Laboratory Head of the
PETER MACCALLUM CANCER CENTRE, VIC VCGS Molecular Genetics Laboratory, one of
Biography the largest clinical laboratories for inherited
Dr Fellowes has worked in the field of disorders in Australasia. She was awarded the
diagnostic molecular genetics in New Zealand Victorian Premier's Commendation Prize for
for over twenty years. At the Molecular achievement in Medical Research in 1998 for
Pathology Laboratory at Canterbury her PhD work, and the Victoria's Young Tall
Health Laboratories, he was involved in the Poppies award from the Australian Institute
establishment of some of the first accredited of Political Science in 1999. She actively
molecular genetics tests in NZ and helped advocates for clinical molecular genetics
create one of New Zealand’s largest clinical within Australasia through roles within the
molecular genetics laboratories which he Human Genetics Society of Australasia and
headed from 2002 until 2011. He completed the RCPA Molecular QAP, and internationally
a PhD in Pathology at Otago University, through roles within the European Quality
which contributed important insights into the Network and the Human Variome Project.
structure and function of the coagulation Abstract
protein fibrinogen. In the last ten years his Next Generation Sequencing [NGS] has
interest has focused on the application of revolutionized molecular testing in genetic
automation in the molecular laboratory as disorders. It has also revealed the complexity
a driver of quality improvement through of gene interactions and how these come
reduced wastage, variation and manual together to produce a clinical phenotype
handling. He has extensive experience in observed in patients. Insight into genetic
robotic systems and was an early implementer variation in the general population is key to
of automated molecular workflow for clinical understanding the contribution of pathogenic
genotyping assays. His current role at the Peter variants to development of the clinical
MacCallum Cancer Centre in Melbourne phenotype observed in patients. This is the
Australia focuses on development of robust essence of establishing the clinical utility of
NGS-based diagnostic assays for cancer NGS technology for use in clinical practice.
diagnosis. Dr Fellowes is a respected member This section of the workshop will discuss what is
of the molecular genetics community in known about human genome variation, how
this information allows the identification of
41
Handbook
2014 AGTA
Conference
pathogenic or potentially pathogenic variants into both lab bench work and informatics
by NGS technology. It will also briefly discuss pipelines. This is further confounded by the
the various applications and designs of NGS reality that many commercial components
and the best fit options for specific clinical care of such pipelines have not been designed
outcomes. strictly with diagnostics in mind nor are
they sold with that intention. This leaves the
laboratory implementing such technologies in
1435 – 1500 a position where they must perform relevant

validation for any assay that is not supplied
as a locally accredited assay for a specific
OPTIMISATION AND VALIDATION application. Key points around the optimization
and validation of NGS technologies in the
Dr Cliff Meldrum diagnostic setting will be presented using
HUNTER NEW ENGLAND HEALTH, NSW examples where possible.
Biography
Dr Cliff Meldrum has recently been appointed
as the Genomics Project Manager for NSW 1500 – 1525
Health Pathology. Cliff has a Fellowship of the
Human Genetics Society of Australasia (FHGSA-
Molecular Genetics) and is a Founding Fellow STANDARDS AND CONTROLS
of the RCPA faculty of Science (FFSc-RCPA).
Dr Natalie Thorne
Cliff’s experience in Molecular Genetics and
MELBOURNE GENOMICS HEALTH ALLIANCE,
Molecular Pathology spans over 18 years, both
AUSTRALIA
within the NSW Pathology Network and also the
Peter MacCallum Cancer Centre in Melbourne Biography
and as an early adopter of Next Generation Natalie Thorne is the clinical bioinformatics and
Sequencing technologies he has overseen their genomics project manager for the Melbourne
validation, accreditation and implementation Genomics Health Alliance. Melbourne
into routine diagnostics. There are a number of Genomics is linking the clinical, research
challenges with these technologies however; and teaching strengths of its seven founding
there are an even greater number of potential members to integrate genomic medicine into
applications likely to dramatically change everyday healthcare for the betterment of
how the analysis of nucleic acids is performed patients.
in diagnostic laboratories. Applications that After a degree majoring in mathematics and
range from targeted regions within genes to statistics and minoring in genetics, Natalie
whole genomes and that encompass many completed her PhD at the Walter and Eliza
of the traditional Pathology disciplines. These Hall Institute’s Bioinformatics division with Prof
applications need to be explored, evaluated Terry Speed and Prof Gordon Smyth (2004).
and where appropriate implemented into She worked for Cancer Research UK at the
routine laboratories. Cambridge Research Institute with Prof Simon
Abstract Tavare in the computational biology group
The implementation of new technologies, no where she worked on a variety of microarray
matter what they are, into clinical diagnostic genomics applications in cancer and analysis
laboratories must be a carefully managed of data from emerging genomic technologies.
process with each step of the process After five years in the UK, she returned to
evaluated and controlled. Next generation the Walter and Eliza Hall Institute in a senior
sequencing technologies are a challenge if for postdoctoral position in the statistical genetics
no other reason than the fact that they are not lab with Dr Bahlo. Her studies utilised high-
a single technology. Rather, each application throughput sequencing technology with
is a collection of processes assembled a focus on cell-free DNA in plasma during
42
Handbook
2014 AGTA
Conference
pregnancy and in urine in the kidney transplant

setting.
1615 – 1640
Abstract
As genomics moves towards integrating into BIOINFORMATICS AND COMPUTING IN THE
the health care system, diagnostic grade SETTING OF CLINICAL GENOMICS SETTING
genomic tests need to be developed. Like all
Dr Mark Cowley
diagnostics, the use of standards and controls
GARVAN INSTITUTE OF MEDICAL RESEARCH, NSW
is crucial to controlling the quality of the test
and assessing its performance to make sure it Biography
is fit for clinical use. I will review what quality Mark Cowley is the head of Translational
control is; discuss the steps in genomic testing Genomics, within the Kinghorn Centre for
that need to be controlled, the resources that Clinical Genomics, and an early career fellow
should be used, what exists so far and what with the Cancer Institute of NSW. His research
else needs to be done. revolves around the clinical applications
of next generation sequencing, for both
diagnostics and discovery, utilising his expertise
1550 – 1615 in bioinformatics and genomics.

Abstract
In this era of clinical genome sequencing,
SAMPLE PREPARATION AND SEQUENCING high quality, reproducible and reliable
bioinformatics is more important than
Professor Graham Taylor
ever. The challenge is to accurately and
UNIVERSITY OF MELBOURNE, VIC
comprehensively identify genetic variants
Biography present in a patient sample, which can
Graham Taylor is the Herman Professor of ultimately be used to diagnose a patient with
Genomic Medicine at the University of a genetic disorder, or determine the clinically
Melbourne, Director of the Victorian Clinical actionable variants in a tumour. This is made
Genetics Service Labroatories and director particularly challenging due to the structure of
of the Australian Human Variome Project. the genome, littered with repetitive elements
Current interests are the development of and pseudogenes, the wide distribution of
diagnostic applications and data pipelines genetic variant sizes, from single bases all the
for Next Generation Sequencing (NGS), way to aneuploidy, the wide range of DNA
including the targeted re-sequencing for amount and quality, from FFPE fine needle
the diagnosis of genetic disease, the use of biopsy all the way through fresh frozen tissue,
NGS is tumour profiling for stratified medicine, and the varied scope of next generation
alignment-free sequence analysis and sequencing tests, from single genes through to
the development of secure means data sharing. whole genomes which usually results in lower
Abstract overall sequencing depth the more genes that
NGS requires template in the form of a library of are targeted.
sequence-able molecules. These are produced
In the context of rare Mendelian genomics,
from purified DNA that may come from a range
most groups are favouring either a targeted
of sources including blood, saliva, fresh, frozen
sequencing approach with large 20-300 gene
or fixed tissue. The source and quality of the DNA
targeted panels (eg cardiomyopathy), or
will influence the choice of both sequencing
whole exome or genome sequencing. The
technology and analysis methods. I will illustrate
bioinformatics analysis begins with extensive
some examples of library preparation methods
data quality control to determine sample
from various types of tissue and discuss their
quality, and to ensure that enough sequencing
effect on the sequence data
data has been generated to meet clinical
specifications. Short sequencing reads (fastq
43
Handbook
2014 AGTA
Conference
files) are often trimmed to discard poor quality high quality variants, which in some cases
bases at the ends of the reads, and then have 100% sensitivity, and as such have been
aligned to the human reference genome. clinically accredited without requiring Sanger
This produces SAM or BAM files which can be sequencing validation.
viewed in most modern genome browsers (eg In this presentation, I will run through the
IGV, UCSC genome browser). Short sequencing bioinformatics and computing aspects of a
reads are usually aligned individually, leading typical clinical genomics experiment, touching
to errors around regions of the genome that on both Mendelian and cancer genomics, and
are challenging to align to (eg known insertions some of the pitfalls.
and deletions), which are typically cleaned
up by realigning all reads around known
indels. Finally single nucleotide and small
insertion and deletion variants are identified 1640 – 1705
through a range of algorithms that identify
recurrent evidence for a variant across multiple
Curation and reporting of genetic and
overlapping sequencing reads. Some groups
genomic data in the modern era (2014!)
are performing copy number variant analysis
on NGS data, mostly based upon the idea of Professor Hamish Scott
read depth, however this is challenging due to CENTRE FOR CANCER BIOLOGY, SA
variability in sequencing depth and the patchy
Biography
sequencing coverage over only targeted NGS
Professor Hamish Scott did his PhD (1992) and
data. Variants are then extensively annotated
first post-doc at the Women’s and Children’s
and ultimately filtered to those that are rare in
Hospital and the University of Adelaide.
healthy individuals, known or predicted to be
During these 7 years he led the discovery of
damaging and associated with the patient’s
genes for 3 rare human diseases. After 11
phenotype. The final challenging task is the
more years, with the persistence of Professor
clinical interpretation, determining which of
John Hopwood and others in academia and
the one or two variants that match the patient
industry, this resulted in either FDA approved
phenotype from a list of often hundreds of
therapy (2003) or clinical trials of novel
variants. Due to the larger scope of these tests,
therapies for these diseases.
variants are usually orthogonally validated,
using Sanger sequencing, prior to return of In 1995, Hamish moved to the University of
results. Geneva Medical School in Switzerland. His focus
was, and remains, the application of genetic
In the context of cancer genomics, in the
and genomic technologies to understand
diagnostic setting, typically the DNA samples
diseases processes to improve diagnoses and
are far more challenging, usually with FFPE
treatment. He led international collaborations
material from either tumour blocks or biopsy,
in identification of human genes causing Down
resulting in low amounts of highly fragmented
syndrome and rare forms of genetic deafness
DNA. This has resulted in many groups adopting
and autoimmunity (e.g. arthritis and multiple
an amplicon based strategy around known
sclerosis). This continues to have profound
cancer hotspot mutations, micropanels of
effects on our understanding of basic biology
single genes (eg BRCA1 and BRCA2) or when
of Down syndrome, hearing and the immune
feasible, across panels of up to 300 cancer
system and lead to new therapeutic strategies
genes. Similar analysis strategies exist to
in these and related diseases. This was also
Mendelian genomics, though amplicon-based
the start of his interest in cancer and leukemia
sequencing data has nuances which must be
as children with Down syndrome have a low
accounted for in the sequence alignment and
incidence of solid tumours and a high incidence
variant calling stages. Typically the sequencing
of leukemia. This is also when he started to work
depth achieved is >500x, which results in very
on familial predisposition to leukemia.
44
Handbook
2014 AGTA
Conference
Hamish relocated to the Walter and Eliza Abstract

Hall Institute of Medical Research (WEHI) in Since the late 1980s, all laboratories involved
Melbourne in 2000 as the Inaugural Nossal in research genetic testing and diagnostic
Leadership Fellow. He was appointed as a genetic testing have always had their own
National Health and Medical Research Council form of ad hoc proprietary databases. In
(NHMRC) senior research fellow in 2001. His the 1990s, the ever increasing numbers of
laboratory identified 2 additional deafness pathogenic mutations and variants of unknown
genes and described the role of a gene in significance (VUS) made it clear that data
reprogramming the DNA of an “adult” cell sharing was required. My own observations
from “normal” to become a gamete (sperm or through that time, and until today, are that
oocyte). He also described a cause of familial many of these attempts at data sharing were
predisposition to leukemia. actually about data control and thus failed.
There was, and is, little reward for contributors
Since returning to Adelaide in January 2008
of data to centralised databases. Diverse data
he has been Deputy and then Head of the
formats of so called locus specific databases in
Department of Molecular Pathology at SA
the early years, was also problematic. This has
Pathology, He is an inaugural member of the
meant that a lot of information with potential
Centre for Cancer Biology an Affiliate Professor
diagnostic significance has not been, and
in both the Schools of Medicine and Molecular
continues to not be deposited to relevant
and Biomedical Science at the University of
databases.
Adelaide and an Adjunct Professor in the
School of Pharmacy and Medical Sciences in It is obvious with the increase in the number
the Division of Health Sciences of the University of sequence variants being described by
of South Australia. He is an NHMRC principal numerous research and diagnostic studies are
research fellow and a Founding Fellow of valuable to share. What are the barriers to
the Faculty of Science (FFSc) of the Royal share genomic information, particularly linked
College of Pathologists of Australasia (RCPA). to phenotypic information? What are the
He recently led the identification of mutations current initiatives to help overcome past and
in a gene in rare families and patients that present problems with data sharing? What is
predisposed them to acute myeloid leukemia the quality of the data in the various resources
(AML), infectious diseases and lymphoedema. available and how should this reflect on out
He has helped develop and introduce reporting of clinical significance of sequence
new technologies and tests for improved variants?
treatment (personalized medicine) and is a
Joint Director of the ACRF Cancer Genome
Facility established at SA Pathology. In these
roles he has been central to introducing both
somatic and germline genotyping using next
generation sequencing to SA Pathology at a
panel, exome and whole genome level.
45
Handbook
2014 AGTA
Conference
accessibility measurements with assays for the

function of genomic elements carrying these
biochemical signatures. We evaluate how
SESSION 6 well different kinds evidence, individually and
in combination, predict Active Chromatin
Elements (ACEs) and the nature of regulation
by those elements. A related challenge is to
TRANSCRIPTOMICS ferret out the regulatory target(s) of any given
cis-acting regulatory element. Using ChIA-
CHAIRS: NICOLE CLOONAN PET data we find that long distance physical
AND NADIA DAVIDSON connection topologies are often complex, with
a given ACE often contacting multiple target
genes and many connection passing over
1500 – 1545 intervening genes. The relationship of these
physical patterns to the specific enhancer
activity of several hundred ACES reveals both
FROM ENCODE BIOCHEMICAL GENOMIC expected and unexpected outcomes.
MAPS TO ENHANCER FUNCTION ASSAYS --
AND BACK: DEFINING THE ACES
Professor Barbara Wold 1545 – 1615
CALTECH, USA
Biography
IDENTIFYING AND CHARACTERISING
Dr Wold is the Bren Professor of molecular
TRANSCRIPTS TARGETED BY SUBNUCLEAR-
biology and Director of the Beckman Institute
BODY ASSOCIATED PROTEINS AND LONG
at Caltech. She began working on genome NONCODING RNAS
structure and gene regulation during embryo
development for her Ph.D. thesis at Caltech, Dr Archa Fox
and developed ways to assay cis-regulatory THE HARRY PERKINS INSTITUTE OF MEDICAL
element function during postdoctoral work at RESEARCH, UWA, WA
Columbia. She established joined the biology Biography
faculty at Caltech in 1981 where she and Research Fellow and Principal Investigator at
her colleagues have focused on learning the Harry Perkins Institute of Medical Research
the architecture and logic of gene networks at the University of WA in Perth. She carried out
that drive cell state transitions. They study her PhD with Merlin Crossley at the University
skeletal muscle development, degeneration of Sydney, where she investigated the
and regeneration as a favored model interaction between zinc-finger transcription
system. Recent work emphasizes new ways factors in hematopoiesis; graduating in
to quantitatively map the inputs and outputs 2000. She then carried out a Post-doctoral
of gene networks in a genome-wide manner Fellowship in Dundee, UK, working with
using “next generation” ultra-high throughput Professor Angus Lamond, investigating how
DNA sequencing, and applying these methods nuclear organization affects gene expression.
to muscle and brain networks. It was during this post-doctoral period that
Abstract she discovered a new subnuclar body termed
This talk will focus on relating complex Œparaspeckles¹ that have formed the basis
biochemical signatures in the genome, of her ongoing research program. In 2006 she
including transcription factor occupancy and returned to Australia with an NHMRC Howard
histone marks, RNA-Seq output, and chromatin Florey Fellowship to establish her laboratory in
WA.
46
Handbook
2014 AGTA
Conference
Abstract Ribonucleoside-Enhanced Crosslinking

Paraspeckles are stress-induced subnuclear and Immunoprecipitation (PAR-CLIP) to
bodies found in mammalian cells that are sequence RNA footprints directly crosslinked
formed by specific nuclear RNA binding to paraspeckle proteins. This approach has
proteins interacting with the long noncoding revealed binding patterns for paraspeckle
RNA (lncRNA) NEAT1. Once formed, proteins across the 23kb NEAT1 RNA, and has
paraspeckles affect gene expression, at least also identified non-NEAT1 RNA targets in the
in part, by the sub-nuclear sequestration transcriptome. Secondly, whole transcriptome
of specific transcription factors. NEAT1 and analyses of CRISPR-engineered NEAT1-/-
paraspeckles are involved in stress responses cultured cells compared to wildtype cells.
such as viral infection and cancer progression. These data indicate that overall NEAT1/
Here, two different types of RNA-seq paraspeckles are responsible for transcriptional
experiments aimed at dissecting paraspeckle activation of target genes, in line with potential
and NEAT1 structure and function will be roles for NEAT1 in cell proliferation through
discussed. Firstly, using Photoactivatable- directly binding to active genes.
47
Handbook
2014 AGTA
Conference
1615 – 1645 1645 – 1700
UNRAVELING DENDRITIC CELL-SUBSET RECURSIVE SPLICING ENABLES COUPLING

COMMITMENT - ONE CELL AT A TIME OF PROMOTER CHOICE TO SPLICING WITHIN
LONG VERTEBRATE GENES
Dr Andreas Schlitzer
SINGAPORE IMMUNOLOGY NETWORK, Dr Christopher Sibley
SINGAPORE UNIVERSITY COLLEGE LONDON, ENGLAND
Sponsored by Fluidigm and Millennium Science Biography
Christopher Sibley is presently a visiting
researcher for a year, working with Dr Mike
Inouye at Melbourne University Department
Abstract
of Pathology. He carried out his DPhil at the
Dendritic cells (DCs) arise in the bone marrow
University of Oxford in the lab of Professor
(BM) from a heterogeneous pool of precursors,
Matthew Wood, before carrying out his post-
which gradually matures from the earliest step
doctoral research with Professor Jernej Ule first
of the macrophage-dendritic cell precursor
at the UK Medical Research Councils Laboratory
(MDP), via the exclusively DC committed
of Molecular Biology in Cambridge, and now
common dendritic cell progenitor (CDP),
at the University College London Institute of
towards migratory DC precursors
Neurology. His work involves the development
termed pre-DCs. Circulating pre-DCs transit and application of functional genomics
via the blood to seed tissues where they approaches to study post-transcriptional RNA
differentiate into two major functionally networks within brain.
specialized DC lineages, CD8α+/CD103+
Abstract
DCs and CD11b+ DCs. However, where and
Recent studies have shown that expression
when progenitor commitment to DC subsets is
of long genes in the mammalian brain can
imprinted during DCpoiesis is currently unclear.
be perturbed by regulatory factors linked to
Precursor populations, defined based on
neurodevelopmental or neurodegenerative
the expression of a limited panel of surface
disorders1-3, suggesting unique regulatory
markers, are likely heterogeneous and part of
mechanisms. Here we find that most genes
a developmental continuum that is subjected
that are long in all vertebrate species are
to “graded commitment” rather than linear
specifically expressed in the brain and contain
development. Thus, in order to gain insights
long first introns with a high incidence of cryptic
into DC precursor heterogeneity, we used
promoters. Using the functional genomics
single cell RNA sequencing to reveal previously
methods of total RNA-seq and cross-linked
unappreciated DC subset specific potential
immune-precipitation (CLIP) we identified newly
within the DC precursor compartment.
discovered recursive splice sites (RSS) in several
Combining single cell sequencing with of the long introns within these genes. These RSS
conventional transcriptomic analysis, CyTOF are highly conserved and are strongly bound
mass cytometry and intra-femoral transfer, we by the spliceosome. We show that recursive
identified for the first time DC subset-specific splicing requires exon definition via a “recursive
precursors as well as previously unknown exon” that is located downstream of the RSS.
checkpoints for DC lineage commitment as Importantly, RSS enable an unusually discrete
early as the CDP stage. regulation of splicing; the recursive exon is not
We have thereby redefined and expanded detectable when the dominant promoter is
the continuum of DC precursors by identifying used, but is completely included when cryptic
DC subset-specific precursors, which will allow promoters are used. Most recursive exons
efficient targeting and manipulation of DC contain premature stop codons, which lead
subsets during health and disease.
48
Handbook
2014 AGTA
Conference
to transcript degradation via NMD. Thus, by if its predicted targets that lie in a common
coordinating inclusion of recursive exons with biological pathway have changed in the
the use of upstream promoters, RSS can control opposite direction. Utilising a database of
the stability of resulting transcripts. Collectively, predicted micro-RNA regulatory targets and
we demonstrate how exon definition and splice pathway annotations, pMimCor circumvents
site competition enable recursive splicing, much of the need to produce large datasets
a new mechanism for dictating promoter- in order to generate stable and accurate
dependent splicing and transcript stability in network estimation. Furthermore, by ranking
the brain. relevant micro-RNA and pathways concurrently
pMimCor also provides highly interpretable
biological information that can be used to
1700 – 1730 infer how a given micro-RNA regulates specific

functional processes.
We demonstrate this framework can
INFERRING DATA-SPECIFIC MICRO-RNA successfully be applied to datasets with
FUNCTION THROUGH THE JOINT RANKING OF small sample sizes, with pMimCor performing
MICRO-RNAS AND PATHWAYS FROM SMALL competitively across a range of performance
SAMPLE MIRNA-SEQ AND MRNA-SEQ DATA metrics when compared to other approaches.
Dr Ellis Patrick When applied to publicly available cancer
UNIVERSITY OF SYDNEY, NSW datasets and our own small Notch2 conditional
knockout experiment with mice, pMimCor
Biography
identifies more micro-RNA that have been
Ellis Patrick is currently a postdoctoral research associated with each specific dataset in the
associate in the School of Mathematics literature.
and Statistics at the University of Sydney. He
identifies as a statistical bioinformatician and
his main research interests are aligned with
both the horizontal and vertical integration of 1715 – 1730
large datasets.
Abstract CROSS-SPECIES SYSTEMS ANALYSIS OF
Understanding how micro-RNA regulate A METABOLIC DISEASE TO FIND A GENE
messenger-RNA is a fundamental biological SIGNATURE PREDICTIVE OF INSULIN
problem. In practice, identifying and RESISTANCE
interpreting the functional impacts of the
Dr Rima Chaudhuri
regulatory relationships between micro-RNA
UNIVERSITY OF SYDNEY, NSW
and messenger-RNA is non-trivial. By leveraging
the information within existing Biography
micro-RNA target and pathway databases Rima completed her PhD in Bioinformatics and
we propose an approach to stabilise the Computational Drug Design in the University
estimation and annotation of micro-RNA of Illinois at Chicago in 2010. Soon after, she
regulation that is suitable for datasets with moved to the Barcelona Supercomputing
small sample sizes. Specifically, we propose a Center and Institute for Research in
supervised framework, built upon concepts of Biomedicine as a postdoctoral candidate in
significance combination, for jointly ranking Spain to work on methodology development.
regulatory micro-RNA and their potential Next, she moved to Sydney, Australia and
functional impacts with respect to a condition joined the Garvan Institute for systems biology
of interest. and bioinformatic research in Diabetes and
Obesity.
Our proposed method, pMimCor, directly tests
if a micro-RNA is differentially expressed and
49
Handbook
2014 AGTA
Conference
Abstract
Insulin resistance (IR) is perhaps the best
predictor of future development of type 2 WEDNESDAY
diabetes (T2D). Identification of molecular 15 October 2014
signatures that can identify individuals with
higher risk of developing T2D could enable
early stage intervention. Gene expression data
could provide an ideal tool for identification
of such a molecular signature, but human SESSION 7
gene expression (GE) data is inherently noisy
and highly variable. To overcome the lack
of power, we integrated human data with
more stable data from a model organism and FUNCTIONAL GENOMICS
thereby developed a cross-species (human-
mouse) analysis platform. We applied this new CHAIRS: KAYLENE SIMPSON
approach to identify a gene signature that
AND AMEE GEORGE
could differentiate insulin resistant from insulin
sensitive individuals with improved prediction
accuracy (~20% increase) compared to 0900 – 0945
standard clinical measures (OCM). This
signature also identified beta-catenin and
JAK1 as novel ‘connection hubs’ to the Insulin EXPLORING IMMUNE SIGNALING SYSTEMS
Signaling Pathway whose GE pattern highly WITH HIGH THROUGHPUT, HIGH CONTENT
correlated to extensive metabolic phenotypic SCREENING
measures of insulin sensitivity in humans.
Dr Iain Fraser
Inhibiting these proteins impaired insulin-
NATIONAL INSTITUTES OF HEALTH, USA
stimulated glucose uptake in L6 myotubes,
confirming their role in insulin action. These Biography
data indicate the potential utility of using Iain Fraser is Chief of the Signaling Systems
systems biology approaches to segregate Unit in the Laboratory of Systems Biology at
individuals early based on differential diabetes the National Institute of Allergy and Infectious
risk. Disease, National Institutes of Health. He
received his B.S. in biochemistry from Heriot-
Watt University, Edinburgh, and his Ph.D.
in biochemistry from Imperial College,
University of London. He was a Wellcome Trust
International postdoctoral fellow at the Vollum
Institute and then co-director of the Alliance
for Cellular Signaling (AfCS) molecular biology
group at the California Institute of Technology.
He was appointed as an Investigator at the
NIH in 2008. His research has focused on the
mechanistic basis of cellular signaling, both in
G protein signaling systems and more recently
in Toll-like receptor signaling in innate immune
cells. He applies systems biology approaches
to decipher how mammalian cells integrate
stimuli in a complex environment to ensure
context-dependent cellular responses. This is
vital to understanding how a breakdown in
50
Handbook
2014 AGTA
Conference
information processing through cell-surface 0945 – 1015

receptors and their linked signal transduction
pathways leads to human disease. Dr.Fraser
has developed sophisticated approaches for SYNTHETIC LETHAL TARGETING OF
RNAi-based perturbation analysis of immune E-CADHERIN-DEFICIENT CANCERS
cell signaling, as well as cloning platforms
Professor Parry Guilford
and plasmid repositories for high-throughput
UNIVERSITY OF OTAGO, NEW ZEALAND
imaging and cell biological applications.
Biography
Abstract
Prof. Parry Guilford is a Principal Investigator
Modern technology now allows the analysis
in the Cancer Genetics Laboratory, University
of immune responses and host-pathogen
of Otago, the Director of the Centre for
interactions at a global level, across scales
Translational Cancer Research, and the Chief
ranging from intracellular signaling networks
Scientific Officer and co-founder of Pacific
to individual cell behavior to the functioning
Edge Ltd (PEL). He completed his MSc at
of a tissue, an organ, and the whole organism.
Otago in 1983, and his PhD at Cambridge
The challenge is not only to collect the large
University in 1989. His current research
amounts of data such methods permit, but
interests include the genetics of inherited
also to organize the information in a manner
and sporadic cancers, in particular stomach
that enhances our understanding of how the
cancer. Other active research involves the
immune system operates or how pathogens
development of genomic-based diagnostic
affect their hosts. Our laboratory is part of a
tools for early cancer detection and the
systems biology program at the NIH whose goal
application of tumour gene expression and
is to develop detailed quantitative models
mutation profiling to the optimised selection of
that can be used to predict the behavior
chemotherapeutics for cancer patients.
of a complex biological system, whose
properties help explain the mechanistic basis Abstract
for physiological and pathological responses Downregulation of the E-cadherin gene (CDH1)
to infection or vaccination and can be used through either mutation or epigenetic silencing
to design new therapies or vaccines. We use is the hallmark of diffuse gastric cancer and
high-throughput genome-scale technologies lobular breast cancer. We propose that this loss
(such as siRNA screening and profiling of creates vulnerabilities in the cancer cell that
RNA expression and protein modification) to can be exploited with drugs (‘synthetic lethal’
broadly characterize the cellular response to interactions). To identify the vulnerabilities
pathogenic stimuli. I will describe data from created by E-cadherin loss, we have
ongoing projects in our lab including: conducted a genome-wide siRNA knockdown
screen, a 4000 compound known drug screen
• Genome-wide RNAi screens to characterize
and a screen of the WEHI’s 114,000 compound
signaling network topology in hematopoietic
WECC library in isogenic MCF10a cell lines
cells and to identify the "parts list" of cellular
with and without E-cadherin expression.
components involved in immune cell responses.
The functional screen has shown that GPCR
• Development of software applications for signalling proteins are highly enriched amongst
comprehensive analysis of siRNA screening the candidate synthetic lethal proteins, as
data. well as many protein classes associated with
• Use of the above methods to characterize microtubule and cystoskeletal function. Drug
the response of immune cells to defined stimuli, classes that show increased activity against
such as single and combined pathogenic the E-cadherin-deficient cells include the JAK
ligands, and intact pathogens, such as inhibitor LY2784544, the tyrosine kinase inhibitor
bacteria and viruses. Crizotinib and several HDAC inhibitors. The
WECC library screen has identified a further 88
novel compounds that show synthetic lethality.
51
Handbook
2014 AGTA
Conference
This finding paves the way for targeted use

of several existing cancer drugs as well as
1030 – 1045
the development of novel drugs and drug
combinations for the chemoprevention and
LONG NON-CODING RNA PATHOBIOLOGY IN
treatment of E-cadherin-negative cancers. COLORECTAL CANCER DEVELOPMENT AND
PROGRESSION
Dr Sheng Liu
1015 – 1030 WALTER AND ELIZA HALL INSTITUTE, VIC
Biography
REPLICATION TIMING IN PROSTATE CANCER Sheng Liu received his PhD in Molecular Biology
from the University of Melbourne in 2013.
Dr Nicola Armstrong
During his PhD, he studied the regulation of
UNIVERSITY OF SYDNEY, NSW
the transforming growth factor-beta (TGF-Î²)
Biography signalling pathway with specific focus on the
After obtaining her PhD in Statistics from UC regulation of TGF-Î² receptors. Since early 2013,
Berkeley, Nicola moved to the Netherlands, he has been a postdoc at the Walter and Eliza
spending time as a post doc before moving to Hall Institute of Medical Research and is running
the Netherlands Cancer Institute in Amsterdam a high-throughput siRNA-based functional study
as a senior statistician. Since returning to on human colorectal cancer cell lines to study
Australia, she spent three years at the Garvan long non-coding RNA pathobiology.
Institute as a senior bioinformatics officer in Abstract
the Epigenetics Program before moving to the Colorectal cancer (CRC) is the second most
University of Sydney where she is a member of common cancer in Australia, affecting 1 in
the Statistics research group. 20 people during their lifetime. In patients
Abstract presenting with non-metastatic disease the
DNA replication, during S-phase of the cell primary cancer can often be resected, but
cycle, is a tightly regulated developmental many individuals remain at risk of recurrence.
process. In a given cell type a specific genomic Outcomes for patients who develop metastatic
locus will always replicate at the same disease are very poor, with a median overall
time point, a quality known as “replication survival of only 24 months, despite an increasing
timing”. In order to investigate if an altered number of chemotherapy and targeted therapy
replication program has associated changes options. LncRNAs are non-protein coding
in the epigenomic landscape, we carried transcripts >200 nucleotides in length that are
out “Repli-seq” on both normal and cancer emerging as important regulators of gene and
prostate cell lines. This technique generates the protein expression in the human genome. To
replication times for all genomic loci. Linking date, several lncRNAs have been implicated in
the replication timing data with epigenetic the development of human colorectal cancer
data on histone modifications and methylation (CRC), but a genome-wide survey of lncRNA
levels, we found that epigenetic gene deregulation with systematic annotation
activation and repression at both individual of function is lacking. We hypothesize that
genes and at domains is associated with a discovery of lncRNAs with critical roles in human
shift in replication timing between normal CRC development and progression can be
and cancer cells. Additionally, genome-wide achieved through integration of genome-wide
hypomethylation, which commonly affects transcriptome analyses on primary cancers with
cancer DNA, was found to occur exclusively high-throughput siRNA-based functional studies
at DNA that replicates late in both normal and on human CRC cell lines. In collaboration with
cancer cells. the Victorian Centre for Functional Genomics
(VCFG), we are the first Australian group to
52
Handbook
2014 AGTA
Conference
gain access to the newly developed Human

Lincode siRNA Library (Dharmacon), which
comprises pools of 4 individual siRNAs against
2,231 validated human lncRNA targets. In a SESSION 8
proof-of-principle study, over 250 lncRNAs
were detected to be differentially expressed in
11 primary cancers, and many of these were
found to impact on CRC cell proliferation SINGLE MOLECULE SEQUENCING
consistently by using a panel of 4 CRC cell lines
in a siRNA-based knock-down screen. CHAIR: DAVID LOVELL
1115 – 1145
RESOLVING GENOMIC AMBIGUITY: INSIGHTS

INTO HUMAN GENOME SEQUENCING AND
PATHOGEN SURVEILLANCE USING VERY LONG
READS
Dr Robert Sebra
ICAHN SCHOOL OF MEDICINE, MOUNT SINAI
INSTITUTE FOR GENOMICS AND MULTISCALE
BIOLOGY, NEW YORK, USA
Sponsored by
Biography
Robert (Bobby) has extensive experience in
using PacBio for human and bacterial genome
sequencing. He has also implemented the
technology in a clinical setting
Abstract
Whether surveying the human genome for
pathologically relevant complex structural
variation or investigating various microbes
related to infectious disease, long read
length sequencing is rapidly enabling more
robust references that include previously
unachievable domains. Recent headway in
single molecule real time (SMRT) sequencing
has enabled faster, more comprehensive
diagnostics assays. Here, we focus specifically
on those methods related to uncovering
genetic structural variation towards a more
comprehensive sequencing pipeline. Despite
the progressive advances in next-generation
sequencing technologies over the past
decade, a variety of hard to sequence
repetitive human genetic loci remain
53
Handbook
2014 AGTA
Conference
ambiguous, including many exonic regions with

expected biological function. We present data
resolving numerous structural variants from
whole-genome sequencing data achieved SESSION 9
using the RSII platform as well as methods for
targeting repetitive genetic regions using long-
range, high fidelity PCR in tandem with SMRT
sequencing. These results demonstrate both a METAGENOMICS
high N50 reference alongside the potential for
diagnostics that better assess genetic variation
CHAIRS: TORSTEN SEEMANN
in pathologically relevant, repeat rich regions
AND MARK SCHULTZ
and that explore the relationship of structurally
complex, non-coding domains using long
reads. Further, these results reinforce the 1345 – 1415
pressing need to derive more comprehensive
and accurate human genome references
to expand their clinical utility beyond that TOWARD ROUTINE CLINICAL SURVEILLANCE
achieved with traditional, shorter read length OF THE HUMAN MICROBIOTA
NGS technologies.
Associate Professor Aaron Darling
Similarly, genetic diversity among infectious UNIVERSITY OF TECHNOLOGY SYDNEY, NSW
disease isolates spans a multitude of unique
Biography
pathological structural and epigenetic variants
Aaron Darling is an Associate Professor in
that identify factors related to virulence, origin,
Computational Genomics and Bioinformatics in
and resistance. Addressing the need to finish
the UTS Faculty of Science's ithree institute. He
microbial genomes to better identify these
has over a decade of experience developing
factors, we present a rapid method to isolate
computational methods for comparative
genomic DNA and generate read lengths with
genomics and evolutionary modeling and in
N50 values as high as 12,000 base pairs using
2013 moved from the University of California-
SMRT sequencing. Through a series of studies,
Davis to start a computational genomics
a demonstration of how readlengths as long as
group at UTS. Darling has led the development
60,000 base pairs have enabled the ability to
of several widely used software packages,
assemble complex structural domains, identify
including the Mauve software for genome
structural insertion elements, phase accessory
alignment & comparison, the mpiBLAST parallel
genomes such as plasmids, and how these
BLAST software, and the BEAGLE library for GPU
reads are used to better understand microbial
accelerated statistical phylogenetics used in
genetics and transmission will be discussed.
BEAST and MrBayes.
One specific example from a solid organ
transplantation case will detail the comparison He is an active advocate for open access,
of Staphylococcus aureus isolates in donor and open source, and open peer review in scientific
recipient and how microbial finishing can be communication.
used to further improve our understanding of Abstract
the epidemiology of bacterial transmission and Bacterial genomes have proven to be
the risk of adverse patient outcomes when a remarkably dynamic entities, containing
compromised organ is considered. Additionally, systems that incorporate foreign gene
better characterization of community isolates content and an ability to maintain substantial
may improve surveillance, reducing the number population-level genetic diversity. Isolates
of hospital-acquired cases and minimizing the of the same species can differ markedly
risk of outbreak. Various microbial case studies in their gene content, with the collection
will be presented to illustrate this development. of genes common to all members of the
54
Handbook
2014 AGTA
Conference
species representing 60% or less of a strain's 1415 – 1445

genome. The extent of variation both in gene
content and point mutations is dependent
on population genetic processes including The infant airway microbiome
population demographics and recombination.
Dr Kathryn Holt
Meanwhile, the human body plays host BIO21 INSTITUTE, UNIVERSITY OF MELBOURNE, VIC
to a complex microbial ecosystem, the
Biography
development of which likely begins before
birth. After birth, the human microbiota Kat completed her PhD in 2009 at the
continues to develop, and emerging Wellcome Trust Sanger Institute and University
evidence suggests that the trajectory of this of Cambridge, studying bacterial comparative
development is associated with long term genomics and genomic epidemiology. In
health outcomes. Routine monitoring of 2010 she returned to Australia to take up an
the development of microbial ecosystems NHMRC ECR Fellowship in the Department
in newborns (or other environments) using of Microbiology and Immunology at the
metagenomic methods is currently extremely University of Melbourne. While working as a
challenging and expensive. I will describe some research fellow, she undertook a Masters in
recent technological advances that could Epidemiology at the university, graduating in
enable routine sequencing and computational 2011. In 2012 she was awarded a Rod Rickards
analysis of hundreds of metagenomes, and Fellowship from the Australian Academy of
demonstrate their application to samples of Science to spend time at the Pasteur Institute
an infant fecal microbiome during the first in Paris, and obtained several NHMRC Project
months of life. In this study forty-five samples Grants to study various aspects of microbial
were subjected to metagenomic sequencing. genomics in relation to human health. In
The microbial community structure in the December 2012 she became a lab head in
resulting data were analysed with a new the Department of Biochemistry and Molecular
approach called phylogenetic Edge Principal Biology at the University of Melbourne. In 2013
Component Analysis (Edge PCA) that can she was awarded a L’Oréal Australia Women
identify which lineages in a phylogeny in Science Fellowship and is now an NHMRC
explain the variation among the samples. Career Development Fellow.
Major life events in the infant appear to be Abstract
associated with dramatic change in the gut Respiratory infections are common in young
microbiota. Strain or substrain-level genomic infants and are a major cause of morbidity
variation is common in many ecosystems and mortality. In the past decade, respiratory
and also appears in the infant gut. This strain- infections during infancy have been
level variation remains especially difficult to recognized as an important factor driving the
characterise. I will discuss the potential for a development of asthma during childhood. In
new technique called metagenomic Hi-C to addition, bacterial colonisation of the airways
help characterise the strain-level variation in has been shown to influence viral infection
microbiome samples. and asthma development. We set out to
characterize the bacterial composition of the
nasopharyngeal microbiome (NPM) in a cohort
of 234 infants at high risk of allergy, which was
previously established to investigate the role of
respiratory infection in asthma development.
The infant NPMs had a simple structure with
six major types and were subject to dynamic
changes during the first year of life, influenced
by childcare, siblings, breastfeeding, season,
antibiotics and incident respiratory infections.
55
Handbook
2014 AGTA
Conference
Early colonization with Streptococcus Abstract

was associated with the development of Identifying susceptibility variants for
allergic asthma by age five. Streptococcus, multifactorial human traits can be complicated
Haemophilus and Moraxella were associated depending on the understanding of the
with the presence and severity of symptoms of genetic background on which it arises. This
respiratory infection, regardless of the presence is especially difficult in small, under-studied
of virus. Moraxella was also a remarkably stable populations without readily-available control
colonizer of the infant airway, whose presence cohorts. In such populations, low-penetrance
was associated with increased incidence and variants in complex disease can be challenging
severity of infections with respiratory syncitial to confidently identify when normal variation
virus (RSV). and allele frequencies are uncharacterised.
We present an investigation of biliary atresia
(BA), a usually sporadic malformation of the
biliary tree. Worldwide, BA leads to half of all
1445 – 1500 paediatric liver transplants, and is always fatal
without major surgical intervention. The causes
MAPPING SHARED MARKERS ACROSS A of BA remain unknown; autoimmune processes
POLYNESIAN POPULATION IN THE GENOMIC and genetic background may both play a
AGE part. In Maori and Polynesian populations the
incidence is elevated three-fold compared to
Ms Sophia Cameron-Christie Europeans. We have identified a large Maori
UNIVERSITY OF OTAGO, NEW ZEALAND family (iwi) exhibiting an extremely elevated
Biography incidence of BA (1:100–300). To circumvent
Sophia Cameron-Christie is a PhD student in some of the problems in studying complex traits
the Clinical Genetics Lab at the University of in Maori we have adopted a non-parametric,
Otago, New Zealand, following an honours family-based approach to localise a
degree in genetics at Otago. She has worked presumptive genetic factor conferring this
to identity mutations for rare, Mendelian susceptibility. Assuming a single, segregating
diseases with next generation sequencing susceptibility factor contributes to BA in our
and to use both array and NGS data for family cohort, we have used the software
linkage analysis and haplotype discovery. Beagle, Germline and in-house methods to
She is interested in studying population history examine haplotype sharing between affected
through genomic data. pairs across this family. We have constructed
maps of long, Identical-By-State (IBS) segments
across the collective genomes of affected
individuals without reliance on pre-existing
assumptions about allele frequencies, a
definitive inheritance model or exclusion of
unidentified phenocopies. Further work aims to
integrate rare variation from next generation
sequencing data to further resolve shared
haplotypes and define candidate regions in
which a susceptibility variant could lie.
56
Handbook
2014 AGTA
Conference
compared to the target sequence with a

1500 – 1515 particular tendency to form gaps (both as
insertions and deletions).Correct and incorrect
calls tended to form clusters along the reads,
PORE PERFORMANCE? BENCHMARKING
suggesting the reads can enter an “error-prone
THE OXFORD NANOPORE TECHNOLOGIES
MINION FOR METAGENOMICS AND CONTIG state” during certain sections or phases of the
ASSEMBLY. read. Both the systematic and random nature
of these errors was also assessed.
Dr Ken McGrath
We assessed the utility of the MinION platform
THE AUSTRALIAN GENOME RESEARCH FACILITY,
QLD in 2 applications where long sequencing reads
are particularly useful – microbial community
Biography profiling and de novo assembly. We trialled
Ken McGrath is the manager of the Brisbane the platform to identify bacteria from a known
Lab of the Australian Genome Research artificial microbial community based on
Facility. He completed his undergraduate sequencing the 16S region, and compared
degree with honours in 2001 at QUT working these results with alternate methods available
with the plant biotechnology group on on established sequencing platforms such
developing transgenic bioreactors, and as the Roche 454 and Illumina MiSeq. We
transitioned to UQ for his PhD work investigating also assessed the utility of MinION reads for
the genetic regulation of plant defence scaffolding contigs in the de novo assembly of
responses to disease, in collaboration short-read sequencing data sets.
with CSIRO and the CRC for Tropical Plant
This work highlights the strengths and
Protection. Following this, his post-doctoral
weaknesses of this emerging technology, and
research with the Schmidt and Schenk labs at
provides an insight into what the future of
UQ involved examining the transcriptomes of
single-molecule sequencing holds.
mixed microbial communities in industrial and
agricultural settings. In 2009, Ken joined the
AGRF as sequencing supervisor, and currently
helps manage submissions and workflows
on a range of next-generation sequencing
platforms.
Abstract
The Oxford Nanopore Technologies (ONT)
MinION is a revolutionary new device capable
of directly sequencing single DNA molecules
via translocation through a protein pore. The
Australian Genome Research Facility (AGRF)
was chosen to take part in the MinION “early
access” sequencing program, which allowed
us to receive and trial the platform prior to
general release.
We have performed several sequencing
runs on this device, including whole-genome
shotgun sequencing, as well as targeted long-
amplicon sequencing.
These initial runs generated between 1000-5000
reads events per 6hr run, with read lengths
ranging from 5bp to over 42,000bp. Analysis
of these reads showed a high error rate when
57
Handbook
2014 AGTA
Conference
POSTERS
58
Handbook
2014 AGTA
Conference
POSTER PRESENTATIONS
MONDAY 13 OCTOBER 2014
# POSTER Title Presenting Author
Targeted Massively Parallel Sequencing of the Mitochondrial

1 Proteome
Hayley Mountford
Modelling Cellular Context to Map Kinase-Substrate Phosphorylation

2 Events
Ralph Patrick
Stemformatics: A user-friendly database of well-curated biological

3 data
Tyrone Chen
Statistical analysis for alleie- and homeolog-specific expression with

4 RNA-seq
Satoru Akama
Introducing Enrich and the Interferome bioinformatics tools, using

5 promoter analysis to understand cytokine signalling pathways
Helen Cumming
6 Analysing Tandem Repeats by High Throughput Sequencing Devika Ganesamoorthy
Bioinformatics approaches, challenges and performance of diversity

7 profiling on three next generation sequencing platforms
Naga Kasinadhuni
The extension of voom to test preferentially spliced events in RNA-seq

8 count data
Charity Law
9 VariantGrid: a graphical interface for a scalable variant database David Lawrence
10 A comparison study of CPU and GPU-based short read alignment tools Xi Li
BRAEMBL Galaxy Connect suite: harnessing the data and compute

11 resources of the powerful BRAEMBL computer cluster through the Webber W.P. Liao
friendly Galaxy interface
COMBINE, a group aimed at students and early-career researchers in

12 bioinformatics, computational biology and systems biology
Andrew Lonsdale
13 In Silico gene prioritisation with BrainGEP Vesna Lukic
14 Controlling the error rate correctly when intersecting gene lists Aaron Lun
STUDENT EARLY CAREER RESEARCHER

59
2014 AGTA
Conference
Integrating Next Generation Sequencing Data with additional

15 ‘omics datasets
Paul Maclean
Risk-conscious correction of batch effects: Maximising information

16 extraction from high-throughput genomic datasets
Yalchin Oytam
17 OncoCis : An annotation tool for cis-regulatory mutations in Cancer Dilmi Perera
DiffVar: A new method to test for differential variability applied to DNA

18 methylation in cancer and aging
Belinda Phipson
Allele Frequency Calibration in Pooled DNA samples: A Machine

19 Learning Approach
Ashfaqur Rahman
Integrative omics analysis elucidates role for inflammation and

20 neutrophil degranulation in glycoprotein-associated all-cause Scott Ritchie
mortality risk
21 Why weight? Variance modelling for designed RNA-seq experiments Matthew Ritchie
Statistical methods and software for functionally characterizing

22 every single mutation in Your Favorite Gene
Alan Rubin
A statistical method to detect pathogenic short-tandem repeat

23 expansions in Massively-Parallel Sequencing
Rick Tankard
24 Optimising multi-caller approaches to somatic variant calling Paul Wang
Differential exon usage detection with RNA-seq using altered count

25 summarisation options
Katrina Bell
Whole-Genome Bisulfite Sequencing (WGBS) An improved “post-bisulfite”

26 conversion library construction method from low gDNA inputs
Leah Alcock
Epigenetic changes in survivors of preterm birth: effect of gestational

27 age and evidence for a long term legacy
Jeffrey Craig
A comparison of control samples for ChIP-seq of histone

28 modifications
Christoffer Flensburg
Ezh2 is essential for lung development, maintenance of epithelial cell

29 identity and cell fate determination
Aliaksei Holik
Genomewide Analysis of Proteins that Bind to DNA and Regulate Gene

30 Expression
Siobhan Hughes

60
2014 AGTA
Conference
31 Epigenetics, T-cells and autoimmunity: A bioinformatics story Jovana Maksimovic
Mutations in LEPR gene are associated with reproductive traits in

32 Davisdale sheep
Michelle French
Exploring the pheromone glands of native New Zealand leafroller

33 moths using de novo transcripts preliminary insights
Amali Thrimawithana
34 A Bovine Gene Expression Atlas Christy Vander Jagt
Comparative transcriptomics of Acacia acuminata parasitised vs. non-

35 parasitised by the hemiparasitic shrub Santalum spicatum
Bjorn Dueholm
Dynamic activities of novel transcripts in leaf and shoot apical

36 meristem of soybean during floral transition
Chol-Hee Jung
Towards Genomic Selection for essential oil yield in Eucalyptus and

37 Melaleuca
David Kainer
Efficient Genotyping-by-Sequencing Enables Construction of the First

38 Linkage Map for the Non-Model Legume Forage Species, Bituminaria Jafar S. Jabbari
bituminosa var. albomarginata (tedera)
Microsatellite marker-based identification and genetic relationships of

39 millennium olive cultivars in Tunisia
Mnasri Sameh
Assembly and analysis of a male sterile rubber tree mitochondrial

40 genome reveals DNA rearrangement events and a novel transcript.
Jeremy Shearman
Single cell-type analysis reveals unique patterns of DNA methylation in

41 the root meristem
Tim Stuart
High through-put Sequencing Guides the Design of DNA Minor Groove Anandhakumar
42 Binding Small Molecule with Improved Binding Specificity Chandran
End-to-end sample quality control for next generation sequencing

43 library preparation and SureSelect Target Enrichment on the Agilent Arunkumar Padmanaban
2200 TapeStation System
Evaluation of Fluidigm SNPtrace™ Panel, with a comparison to DNA

44 QUAL and bioanalyzer fragment analysis, for DNA quality assurance Li Zhou
within the Tumour Bank program
45 A Better Way - AmpliSeq Exome for Clinical Research Richard Allcock

61
2014 AGTA
Conference
POSTER PRESENTATIONS
TUESDAY 14 OCTOBER 2014
Exomic molecular signatures segregate oral epithelial lesions based

46 on disease severity
Nigel Bennett
Genomic evolution of hepatocytes towards hepatocellular

47 carcinoma
Magdalena Budzinska
Detecting tumour DNA in blood plasma using allelic imbalance from

48 read countdata
Wenhan Chen
The role of MEK signaling in driving mesenchymal transition in cell line

49 models of human breast cancer metastasis
Erik Thompson
50 Deciphering genomic evolution in heterogeneous tumours Vincent Corbin
miRNA-target gene relationship analyses identify key oncogenes for

51 Kidney Renal||Clear Cell Carcinoma progression
James Doecke
Amplicon-based targeted resequencing identifies novel genetic risk

52 factors for sarcoma and an association between genetic load and David Goode
age of onset in sarcoma patients
53 MicroRNA prognostic signatures in metastatic melanoma Kaushala Jayawardana
Next generation sequencing of haematological diseases in a clinical

54 setting a pilot study
Anabel Kearney
Molecular profile of oral mucosal lesions characterised by direct

55 tissue autofluorescence
Farzaneh Kordbacheh
Exome sequencing reveals Bcl6 co-repressor (Bcor) as a frequently

56 mutated tumor suppressor gene in Eμ-Myc lymphoma
Marcus Lefebure
GENOME-WIDE MULTIMARKER MODEL TO PREDICT THE RISK OF BLADDER CANCER Evangelina Lopez De
57 RECURRENCE Maturana
Stratification of the cancer genomic landscape through the

58 development of a pan-cancer genetic screening panel
Mark McCabe
59 Structural Variation in Pancreatic Cancer Genomes Kelly Quek

62
2014 AGTA
Conference
60 Colorectal cancer methylation biomarkers Jason Ross
Selection Rather Than Mutation Drives Increased Platinum Resistance in a

61 Patient Derived Xenograft Model of Ovarian Cancer
Matthew Wakefield
Whole exome sequencing identifies a recurrent RQCD1 P131L mutation in

62 cutaneous melanoma
Stephen Wong
63 Implementing Clinical Genomics for diagnostics and research Mark Cowley
Genome characterisation of the cell line U937 provides further

64 evidence for centromere capture and demonstrates the use of multiple Ruth MacKinnon
genome technologies to examine evolution in cell lines
Library preparation, targeted-hybridisation capture and sequencing of

65 limiting amounts of FFPE samples
Gisela Mir
66 A clinical ready system for scalable genomics analysis Mahtab Mirmomeni
67 Next-generation Sequencing - Beyond Variants John Pearson
Challenges to CNV Detection in Clinical Settings using Targeted High

68 Throughput Sequencing Data
Simon Sadedin
69 A Handheld qPCR Device for Point-of-Care and In-Field Diagnostics Jo-Ann Stanton
Generating Actionable Knowledge from Complex Genomic Data for

70 Personalised Clinical Decisions.
Siamak Tafavogh
71 Next generation sequencing for pharmacogenetic screening Kim Ton
72 Causes of Multiple Sclerosis: a functional genomics approach Margaret Jordan
How does chromosome 21 trisomy alter the gene regulatory network

73 of neural differentiation?
Nicholas Matigian
74 Epilepsy: new genes and old pathways in a complex disorder Leanne Dibbens
Cross-species Systems Analysis of a Metabolic Disease to Find A Gene

75 Signature Predictive of Insulin Resistance
Rima Chaudhuri

63
2014 AGTA
Conference
76 Imperfect centered miRNA binding sites are common and functional Nicole Cloonan
77 Deriving Mycobacterium tuberculosis Spoligotypes from NGS Data Thomas Conway
78 Genotyping microsatellites in next-generation sequencing data Harriet Dashnow
Identifying disease-associated pathways in genome-wide association

79 studies via a network-based kernel approach
Saskia Freytag
Gene discovery in large multiplex families with autism spectrum

80 disorders
Miriam Fanjul Fernandez
The genetic basis of Leigh syndrome, the most common mitochondrial

81 disorder affecting children: A cohort study
Nicole Lake
Computational deconvolution of genome wide expression data from

82 Parkinson’s and Huntington’s disease tissues using Population-Specific Victoria Perreau
Expression Analysis.
MUTATIONS IN THE mtorc1 REGULATOR DEPDC5 ARE A MAJOR CAUSE OF

83 LESIONAL AND NON-LESIONAL FOCAL EPILEPSY
Michael Ricos
84 Inferring Identity by Descent on the X Chromosome Lyndal Henden
Genotyping-by-Sequencing (GBS): Comparison to SNP chips and Whole

85 Genome Sequencing
Tracey Van Stijn
Extensive genetic co-regulation drives highly correlated modular

86 structures in transcriptomic data
Anita Goldinger
The discovery of novel lincRNA with potential roles in mESC

87 development
Beth Signal
Diversification of innate immune responses by transcriptional

88 mechanisms
Suzanne Butcher
Complex coding and noncoding transcriptional dynamics in early

89 stem cell development
Brian Gloss
90 Detecting Differential Exon Usage in RNA-Seq Yunshun Chen

64
2014 AGTA
Conference
Characterizing changes in epigenome and transcriptome during

91 mammary stem cell differentiation
Wei Shi
RNA-Seq analysis of lip tissue from a mouse model of Williams-Beuren

92 Syndrome
Susan Corley
Discovery of allele specific expression in RNAseq data from eighteen

93 bovine tissues
Amanda Chamberlain
94 Stranded versus non-stranded RNA-seq Sonika Tyagi
Reagents - Consumables - Instrumentation - Solutions
The Latest Advancement in Single-Cell RNA Seq

Pushing the limits of sensitivity for single cells.
REAGENTS - CONSUMABLES - INSTRUMENTATION - SOLUTIONS
The Clontech® SMARTer® UltraTM Low Input RNA Kit for Sequencing - v3
cDNA synthesis for single-cell transcriptome analysis that delivers:
1. Greater number of genes identified

2. Better exon to intron ratio, higher representation of GC-rich genes
and less than 2% RNA reads
3. Full-length gene information from 1-1,000 intact cells or as little
as 10 pg-10ng of total RNA
4. Compatible platforms: Illumina®, Ion Torrent
Scientifix Pty Ltd. 1800 007 900 info@scientifix.com.au www.scientifix.com.au
Illumia is a regestered trademark of Illumia, Inc. For Research Use Only. Not for use in diagnostic or therapeutic procedures. Not for
resale. Your use of these products and technologoies is subject to compliance with any applicable licensing requirements described
on the product’s webpage at www.clontech.com. It is your responsibility to review, understand and adhere to any restrictions
imposed by such statements. Clontech, the Clontech logo, SMARTer and Ultra are trademarks of Clontech Laboratories, Inc. All
other marks are the property of their respective owners. Certain trademarks may not be registered in all jurisdictions.
Clontech is a Takara Bio Company © 2014.

65
2014 AGTA
Conference
Monday purifies with binding partner UQCC1 and have

proven that both are required for early CIII
13 October assembly (Tucker et al, PLoS Genet 2013).
MitoExome sequencing shows promise as a
research tool for discovering novel disease loci
Poster 1 and as a noninvasive diagnostic approach for
patients with clinical evidence of mitochondrial
disease, however proof of pathogenicity still
Hayley Mountford , Alison Compton , Elena remains challenging.
Tucker, Sarah Calvo , Bas Wanschers , Radek
Szklarczyk , Michael Ryan , Martijn Huynen ,
Vamsi Mootha , David Thorburn Poster 2
Targeted Massively Parallel Sequencing of
the Mitochondrial Proteome
Ralph Patrick, Kim-Anh Le Cao, Bostjan Kobe,
Abstract Mikael Boden
Inherited defects in mitochondrial oxidative
Modelling Cellular Context to Map Kinase-
phosphorylation (OXPHOS) are the most Substrate Phosphorylation Events
common inborn error of metabolism, affecting
at least 1 in 5000 live births, predominantly Abstract
affecting organs with high-energy The determinants of kinase-substrate binding
consumption. OXPHOS diseases are notoriously can be found both in the substrate sequence,
difficult to diagnose, as they show extreme and the surrounding cellular context. Cell
clinical heterogeneity, presenting at any age cycle progression, interactions with mediating
and with any level of severity, and typically proteins and even prior phosphorylation events
impact on multiple organ systems. are necessary for kinases to maintain substrate
specificity. While much work has focussed on
They are also genetically heterogeneous with
the use of sequence based methods to predict
over 140 mitochondrial DNA and nuclear DNA
kinase-specific phosphorylation sites, lack of
encoding genes associated with OXPHOS
specificity in many kinase binding motifs means
disease. Despite this, ~50% patients with
that sequence methods for predicting kinase
OXPHOS disease do not have a molecular
binding sites are susceptible to high false-
diagnosis.
positive rates. Few studies have attempted
We developed a targeted DNA capture and to leverage the cellular context that kinases
massively parallel sequencing method to operate in; and while context information
detect variants within 1034 genes encoding is readily available in various databases,
the mitochondrial proteome (MitoExome). We incomplete coverage and variable certainty
applied this MitoExome sequencing approach means that the integration of context features
to a group of 44 unrelated patients with into a model is non-trivial.
clinical and biochemical evidence of severe
We present a Bayesian network model that
OXPHOS disease. Molecular diagnoses were
can accommodate missing values, seamless
provided for 13 patients with mutations in 10
combination of heterogeneous data, and to
known OXPHOS disease genes. This approach
provide flexible options for querying potential
identified 7 novel confirmed disease genes in 9
kinase substrates. The model integrates
patients; NDUFB3, AGK, MTFMT, EARS2, LYRM4,
known kinase-substrate relationships, protein-
COA6 and UQCC2.
protein interactions, and protein abundance
We identified UQCC2 as a novel disease gene across the cell cycle to predict kinase
in a consanguineous Lebanese patient with substrates across a variety of human kinase
severe CIII deficiency, and demonstrated families. Our model shows high prediction
aberrant splicing. We showed that UQCC2 co- accuracy, with a mean AUC of 0.86 across
66
Handbook
2014 AGTA
Conference
the kinases tested. When using the model to generation sequencing, proteomics and
complement sequence based kinase-specific ChIP-seq platforms. Currently the database
phosphorylation site prediction, we find contains mouse and human expression data,
that the additional information can greatly with support for additional model organisms
increase prediction performance at low false being planned. The user interface is designed
positive levels. Our results demonstrate that for rapid assessment of the expression of
the model can account for the short-falls in individual or multiple genes across and/or
sequence information and provide a robust within datasets. Users can store gene lists and
description of the regulation of kinase-protein share views with collaborators. Stemformatics
phosphorylation. is funded as part of the ARC special initiative
"Stem Cells Australia", has an international
community of users and is freely accessible to
Poster 3 all academic groups.
Mr Tyrone Chen, Mr Rowland Mosbergen, Mr Poster 4

Othmar Korn, Associate Professor Christine Wells
Stemformatics: A user-friendly database of

Satoru Akama, Rie Shimizu-Inatsugi, Kentaro
well-curated biological data
Shimizu, Jun Sese
Abstract
Statistical analysis for alleie- and
The exponential growth of 'big-data' to address
homeolog-specific expression with RNA-seq
biological networks has greatly increased the
type and scale of raw data available to the Abstract
research community. This led to the creation of Next-generation sequencers allow us to
publicly-accessible computational repositories distinguish expression levels of gene pair in
for biological data, allowing for external re- homologous chromosomes, such as paternal-
analyses and further studies on datasets. and maternal-specific expressions in diploid
However, existing databases are driven by species or homeolog-specific in allopolyploid.
data-generators, not data-users. This means Recent researches show that the expression
that the quality of the experimental paradigm balances are greatly changed according to
and data is uncurated, and needs to be cellular status, causing differentiations and
examined dataset-by-dataset. Data formats adaptations in various species. However,
vary considerably between platforms, which the lack of its statistical assessment method
coupled with their volume makes it difficult has prohibited the connection between the
to identify datasets relevant to any specific expression difference of the gene pair and
biological question. the understanding of biological functions.
To address this, we have developed a For example, while existing research uses
biology-centric approach. Stemformatics Fisher’s exact test or chi-square test to check
was designed with the aim of integrating the changes of expression balance, our
biology, computational methods and statistics computational experiments confirmed the
in a format intuitively accessible by stem high false positive ratio of the detection
cell researchers. The Stemformatics project with traditional statistics because of their
currently houses 236 datasets, of which 122 no consideration of replicate experiments.
are publicly available. Each dataset has To solve the problem, we introduce a new
been carefully curated for cell phenotype statistical analysis procedure for allele- or
and experimental design, with rigorous homeolog-specific expressions including two
quality control steps. Available datasets new methods. One is to detect the genes
comprise gene expression data from a range whose expression ratios are changed between
of platforms including microarrays, next- different environments named HomeoRoq
67
Handbook
2014 AGTA
Conference
(http://seselab.org/homeoroq/) [Akama et Methods

al., NAR, 42(6):e40, 2014]. The other is the The Interferome database was queried for
statistically sound detection method of allele- genes that were up-regulated in murine cells
or homeolog-specific silencing based on the over discrete time periods following stimulation
estimation of the distribution of genome- by type I Interferons. Transcription factors that
wide expression ratio using a mixture model. were significantly enriched (p<0.05) among
We applied the procedure to allotetraploid each gene set were identified using the Enrich
Arabidopsis kamchatica derived from diploid custom bioinformatics tool.
A. lyrata and A. halleri to examine the
Results
environmental adaptation in evolutionary
process. The results showed that the procedure Interrogation of the Interferome database
could reveal that homeolog sensitive genes identified hundreds of genes that are up-
were highly related to the stress response regulated following stimulation by Interferon.
genes, contributing to the environmental Analysis of the associated promoters using
adaptation. Enrich, identified clear patterns in the
transcription factors binding sites associated
with temporally co-expressed genes. Many
Poster 5 transcription factors identified are already
known to be involved in the Interferon
pathway, including STATs and IRFs. The analysis
Dr Helen Cumming, Dr Sam Forster, Dr Alex also revealed the additional transcription
Finkel, Dr Ross Chapman, Dr Jamie Geering, Dr factors which might provide new insights into
Kevin Luu, Mr David Hughes, Prof. Paul Hertzog Interferon signalling.
Introducing Enrich and the Interferome Conclusions

bioinformatics tools, using promoter By integrating the Interferome and Enrich
analysis to understand cytokine signalling
bioinformatics resources, it has been possible to
pathways
explore the transcription factors that regulate
Abstract different phases of the interferon response
Introduction and identify new regulatory elements for
The interferon cytokine family plays a pivotal experimental validation.
role in protecting the body from pathogenic
infections and tumour development. The
interferon signalling pathway is a tightly
regulated and highly structured temporal
response, with stimulation activating an
organised sequence of transcriptomic events.
Here we present two bioinformatics tools that
can be used to analyse the transcriptional
regulation of the Interferon response: 1) the
Interferome v2.0 database (http://interferome.
its.monash.edu.au/interferome/home.jspx),
which identifies genes regulated by Interferons,
and 2) the Enrich custom software, which is
a powerful and intuitive tool for identifying
transcription factors involved in regulating
co-expressed gene sets and thus elucidating
signaling pathways.
68
Handbook
2014 AGTA
Conference
Poster 6 Poster 7
Dr Devika Ganesamoorthy, Dr Minh Duc Cao, Mr Naga Kasinadhuni, Mr Lavinia Gordon

Mr Wenhan Chen, Dr Lachlan Coin
Bioinformatics approaches, challenges
Analysing Tandem Repeats by High and performance of diversity profiling
Throughput Sequencing on three next generation sequencing
platforms
Abstract
Abstract
The contribution of genetic variation to
complex disease susceptibility has been High-throughput sequencing is reshaping the
extensively studied in the recent years. landscape of microbial studies, revealing the
Although the influence of copy number hidden secrets of the “uncultured microbial
variations and SNPs on complex diseases have world”. Diversity profiling has enhanced the
been explored in detail, the effect of tandem discovery and profiling of microbial species
repeats (TRs) are poorly understood. There are that cannot be isolated into a pure culture,
almost 1 million TRs in the human genome, by sequencing molecular markers such as
however only few of these regions have been the conserved and hypervariable regions
investigated in terms of disease association. of ribosomal RNA. Most samples consist
This is mainly due to the hyper-variable and of an extremely heterogeneous microbial
multi-allelic nature of TRs which makes it community and may contain thousands of
difficult to be tagged by SNPs for genome species that vary significantly in proportion
wide association studies. The recent advances from one species to the next. Sequencing
in high throughput sequencing (HTS) provides these samples using high-throughput platforms
an opportunity to unlock this type of variation. outputs an enormous amount of data, ranging
We have developed a target enrichment from 10^5 to 10^9 reads per run, with each
based sequencing method to analyse 140 platform producing different read lengths,
TRs associated with obesity. The targeted quality, error rates and data formats. Despite
TRs range from 2bp to 1800bp in repeat unit this abundance of data, deciphering some
length and varies from 2 to 2300 number of meaningful information presents a serious
copies. A pilot study was performed to assess challenge to bioinformaticians. From QC to
the feasibility of the targeted sequencing classifying the sequenced amplicons, a series
approach. More than 90% of sequence reads of various complex computation methods are
mapped to the targeted regions, which required. The Australian Genome Research
includes the TRs and the flanking regions. Facility (AGRF) currently operates a microbial
Greater than 100X coverage was achieved in diversity profiling service targeting various
70% of the targeted regions and notably, at regions of 16S, ITS and 18S microbial genes.
least 50X coverage was achieved in 60% of To compare and evaluate the accuracy
the targeted TRs. Longer sequence reads and and reproducibility of various platforms,
optimized read alignment methods facilitated AGRF pooled both artificial control samples
effective analysis of TRs and these findings and environmental samples, and sequences
were independently validated by fragment using GSFLX, MiSeq and Ion Torrent NGS
analysis method. The target enrichment based platforms. The resulting data was analysed
sequencing of TRs provides new opportunities and classified at different taxonomic levels
to explore the impact of genetic variation on using various bioinformatics pipelines. Here we
complex diseases. present the pipelines and results comparing the
three NGS platforms.
69
Handbook
2014 AGTA
Conference
Poster 8 Poster 9
Charity W Law, Katarina Matthes, Malgorzata Mr David Lawrence, Dr Andreas Schreiber, Dr

Nowicka, Yifang Hu, Yang Liao, Gordon K Jinghua Feng, Mr Joel Geoghegan
Smyth and Mark D Robinson
VariantGrid: a graphical interface for a
The extension of voom to test preferentially scalable variant database
spliced events in RNA-seq count data
Abstract
Abstract Sequencing and variant calling is an
In RNA-sequencing, methods for the detection increasingly popular means of detecting
of changes in isoform usage can be classified individual genetic variation, promising to
into two broad categories: counting (e.g., deliver understanding, diagnosis and treatment
exon-level) reads according to existing of many genetic diseases such as inherited
annotation catalogs, or assembling transcripts disorders and cancer.
and estimating their abundances. Either way, Extracting meaning or a diagnosis from
these are used in downstream statistical tests this data has proven difficult, requiring an
to call changes in isoform preference resulting expert knowledge of phenotype, genes and
from alternative splicing. Current methods for pathways while simultaneously managing very
differential splicing (DS) analyses can be further large datasets. These skills are often not found
improved both statistically and in terms of together, slowing progress. We have built a new
computational expense. user interface specifically designed to allow
Unlike other existing methods that follow non-programming experts to search through
discrete probability distributions, we present millions of variants.
an extension to voom (mean-variance Our drag & drop interface allows users to
modeling at the observational level) which, create a Directed Acyclic Graph (DAG) of
after application of precision weights, performs variant filtering and calculation operations.
DS analyses under a normal distribution. The Nodes change to reflect configuration and
method, voom-diffSplice, tests for preferentially results, allowing a high level view of the
spliced transcripts at both the exon- and analysis. These can be connected together to
gene-level. Using both simulated and real build custom analyses, which are converted
data, we show that voom-diffSplice performs into efficient queries and run on large
favorably against existing methods in terms of databases.
error control, it produces biologically relevant
results, and only takes a small fraction of the
computation time required by other existing
methods.
70
Handbook
2014 AGTA
Conference
Poster 10 Poster 11
Dr Xi Li, Mr Joel Ludbey-Bruhwel, Mr Ondrej Dr Webber W.P. Liao, Mr Eric H. Powell, Mr

Hlinka, Dr Josh Bowden, Dr Tim Ho, Dr Annette Alexander Varlakov, Mr Gavin D. Graham
McGrath
BRAEMBL Galaxy Connect suite: harnessing
A comparison study of CPU and GPU-based the data and compute resources of the
short read alignment tools powerful BRAEMBL computer cluster through
the friendly Galaxy interface
Abstract
Abstract
The development of efficient short-read
alignment tools has still been considered as From detecting sequence contamination to
a challenging topic in Bioinformatics due inferring protein homology, BLAST is applied in
to the largely accumulated short sequence many aspects of bioscience research. While
reads generated from the high-throughput but queries of a single sequence can be performed
low-cost next generation sequencing (NGS) using the NCBI or EBI websites, queries with
platforms. Since last few years, a number of hundreds of sequences or more are not so
graphic processing units (GPU)-based short- straight-forward. One may use the programmatic
read aligners have been proposed to fully access provided by websites or set up a
utilize the parallel power of GPUs (e.g., SOAP3, BLAST server. However, both require technical
Barracuda, and PEANUT), which aim to further knowledge and are not suitable for biologists
accelerate the alignment process without the who are uncomfortable with programming and
loss of accuracy in contrast to the CPU ones. configuring systems.
Especially, NVidia has implemented bowtie Galaxy is a web-based platform for bioinformatics
(NVbowtie) for GPUs so that it affords the tools. Its friendly interface allows even a novice
opportunity to make a direct comparison of user to perform simple tasks. Despite this, the
tool’s performance under CPU vs GPU. Here, existing plugins for BLAST still require configuring
we have carried out a comparative study to the tool and maintaining the necessary databases,
investigate the capabilities of GPUs and GPU- which is not trivial for non-bioinformaticians.
aware short-read aligners against CPU ones To relieve users from the tedious configuration and
to understand in what circumstances it would maintenance, we have developed a new plugin
be beneficial to use GPUs and what limitations (BRAEMBL Galaxy Connect) that harnesses the data
there are in our current GPU infrastructure. The and compute resources of Bioinformatics Resource
main facets to examine are speed, accuracy EMBL Australia (BRAEMBL) and makes them
and limitations in terms of data size. This accessible through Galaxy. Once queries are
presentation outlines the datasets and tools we constructed within Galaxy, they are sent to the
chose, the experimental workflows, preliminary BRAEMBL server for processing, leaving the Galaxy
outcomes and related experience we have server’s computing power for other tasks. Results
learnt. are presented in a standard BLAST report as well
as a tabular format for downstream analyses. The
underlying JDispatcher system also generates
visualisations of the results (when available),
which are included in an HTML output for easy
interpretation.
The cloud-based architecture of “BRAEMBL
Galaxy Connect” offers an intriguing alternative
to traditional high-performance computing and
shows promise for addressing other computational
challenges in the field of bioinformatics.
71
Handbook
2014 AGTA
Conference
Poster 12 Poster 13
Mr Andrew Lonsdale Ms Vesna Lukic, Ms Karen L Oliver, Ms Ashley

Marsh, Dr Natalie Thorne, Professor Ingrid E
COMBINE, a group aimed at students and
Scheffer, Professor Samuel Berkovic, Ms Kate
early-career researchers in bioinformatics,
computational biology and systems Pope, Associate Professor Paul J Lockhart, Dr
biology Richard J Leventer, Associate Professor Melanie
Bahlo
Abstract
COMBINE is a student-run Australian In Silico gene prioritisation with BrainGEP
organisation for researchers in computational
Abstract
biology, bioinformatics, and systems biology.
BrainGEP is an R package that makes use of
COMBINE is the official International Society for
developing and adult Allen Human Brain Atlas
Computational Biology (ISCB) Regional Student
data to perform in silico gene prioritisation
Group (RSG) for Australia.
for neurological diseases. Given a set of
We aim to bring together students and early- reference genes (genes previously identified as
career researchers from the computational containing disease causing variants), BrainGEP
and life sciences for networking, collaboration, will calculate pairwise gene correlations
and professional development. using Pearson’s and Spearman’s correlation
Australia has many research institutes, each measures using a weighted sum approach with
with their own cohorts of students. Aside from weightings derived from the sample variances
conferences, there are few opportunities that of the individual pairwise correlation measures.
bring these students together, allowing them to These correlations are then compared to
discover the different kinds of research going a set of randomly chosen genes, using the
on at other institutes. empirical cumulative distribution function. The
reference gene correlations can also be used
COMBINE aims to bridge this institutional divide
to construct networks and correlation plots as
by organising seminars, workshops and social
additional ways to gauge the strength of the
events, also as the yearly COMBINE Student
reference gene set. The reference genes can
Symposium. Together, these events allow
be used to prioritise a set of candidate genes
students to connect with each other and build
by calculating the connectivity between each
a network in a casual environment.
candidate gene and the set of reference genes.
We show results from a recently published paper
by Oliver et al in which a set of 179 candidate
Epileptic Encephalopathy (EE) genes were
prioritised according to their connectivity with
a set of 29 genes known to cause EE. Two of
the highest-ranking genes candidate genes
prioritised by BrainGEP have subsequently been
confirmed as true EE genes. In a second study of
a consanguineous family affected with Agenesis
of the Corpus Callosum (AgCC), BrainGEP
ranked the gene encoding the causal mutation
the highest, following linkage filtering. These
results demonstrate the benefit of incorporating
BrainGEP in gene discovery efforts, as a proven
in silico prioritisation mechanism for brain
diseases.
72
Handbook
2014 AGTA
Conference
Poster 14 Poster 15
Mr Aaron Lun, Professor Gordon Smyth Paul Maclean, Vanessa Cave, Marlon Reis and
Christine Couldrey
Controlling the error rate correctly when
intersecting gene lists Integrating Next Generation Sequencing
Data with additional ‘omics datasets
Abstract
Intersection of lists of differentially expressed Abstract
(DE) genes is often performed to identify genes It is becoming more and more common to
that are changing across multiple comparisons. collect data from multiple ‘omics sources on
Knowing that a gene is potentially involved in an individual sample. Exploration of ‘omics
multiple biological processes can provide more data can shed light on the molecular factors
insights into its function. However, as with all that help determine the phenotype of an
statistical procedures, a degree of uncertainty individual, and a greater understanding of the
is involved in the intersection operation. This underlying biological system can be achieved
uncertainty is represented by the type I error by simultaneously looking at data from multiple
rate of the intersection, i.e., the probability of sources.
obtaining a gene in the intersected subset that This talk provides a brief overview of integrating
is actually non-DE in one or more comparisons. Next Generation Sequencing data with
A common strategy is to identify DE genes additional ‘omics datasets, and highlights some
in each comparison at a particular false methodology and considerations for a new
discovery rate (FDR), and then identify the biology.
shared subset of genes that are DE in each The talk is centred on research on livestock at
comparison. This approach will not control AgResearch Ltd.
the type I error rate, nor will it control the
corresponding FDR in the intersected subset.
Proper control of the type I error rate should
instead be maintained with an intersection-
union test (IUT). This is a very conservative test
in most cases, and will result in loss of detection
power.
Here, the inappropriateness of FDR-based
intersections and the need for the IUT are
demonstrated with some examples. A more
powerful version of the IUT is also proposed for
cases involving two DE comparisons that are
technically independent, i.e., using separate
datasets. The performance of this new method
is tested on both simulated and real data.
73
Handbook
2014 AGTA
Conference
Poster 16 Poster 17
Dr Yalchin Oytam, Dr Fariborz Sobhanmanesh, Dilmi Perera, Diego Chacon, Julie A. I. Thoms,
Dr Jason Ross, Dr Konsta Duesing Rebecca C. Poulos, Dominik Beck, Peter J.
Campbell, John E. Pimanda and Jason W. H.
Risk-conscious correction of batch
Wong,
effects: Maximising information extraction
from high-throughput genomic datasets OncoCis : An annotation tool for cis-
regulatory mutations in Cancer
Abstract
Identification and effective removal of batch Abstract
effects is a major challenge in genomic Whole genomes are being sequenced at
datasets. Batch effects are structured technical an accelerated pace but annotation of
noise which persist under best laboratory mutations and inference of their functional
practice, and typically account for more of the significance remains challenging. Whilst a
data variance than any of the experimental myriad of tools are available for the annotation
factors. Even with ideal experimental design, of protein coding mutations, few are suited
they artificially inflate within-group variances, for annotating non-coding mutations. To this
significantly reducing the power of statistical end, we have developed, OncoCis, to provide
tests, and resulting in actual effects going an easy to use web service for researchers to
undetected. Conversely, if “overcorrected” annotate and prioritize potential causal cis-
– i.e. genuine biological variance is removed regulatory mutations. In order to annotate
along with batch noise – the result is an artificial non-coding mutations, OncoCis integrates
deflation of within-group variances and as publicly available datasets from genome-
a consequence, false positives. Here, we wide DNase-seq and histone ChIP-seq data
present a novel technique which maximises the obtained from ENCODE and the Epigenome
removal of batch effects, with the constraint Atlas, to identify mutations that occur within
that the probability of overcorrection is kept potential cis-regulatory regions in a cell type
to a fraction which is set by the end-user. specific manner. These mutations are further
We benchmark the new technique against annotated with sequence conservation scores
the leading technique currently used, using and searched for possible removal or creation
data from published studies, with the two of transcription factor consensus binding
performance criteria being removal of (batch) motifs. Finally, FANTOM 5 data combined with
noise using guided-PCA as measure (Reese the GREAT tool is used to map mutations to
et. al, 2013), and preservation of (biological) the most likely gene that it may be regulating.
signal. The new method performs better If gene expression data is available, a fold
than the leading method on both fronts change is calculated for each of the mapped
simultaneously. For all datasets, while both genes. We have applied this method to
methods appear to remove batch noise whole genome sequencing data from 21
(guided-PCA p-values >> .05) our method breast cancer samples. Eighteen putative
does so with markedly lower probabilities for cis-regulatory mutations were identified
batch effect persistence, and manages to do in a number of samples that significantly
so while removing significantly less variance increased the expression of the mapped gene
from the datasets. Evidence suggests that when compared with other samples without
the novel method for batch correction, the mutation.This tool will be invaluable to
applicable to microarray and sequencing data researchers seeking to prioritise large lists
alike, is a notable advance on techniques of mutations for evaluating their functional
currently used. A software package has been relevance.transcription factor consensus
developed for public release. binding motifs. Finally, FANTOM 5 data
74
Handbook
2014 AGTA
Conference
combined with the GREAT tool is used to map We applied DiffVar to several publicly available
mutations to the most likely gene that it may be cancer datasets from The Cancer Genome
regulating. If gene expression data is available, Atlas, as well as a publicly available aging
a fold change is calculated for each of the dataset, all of which had samples profiled using
mapped genes. We have applied this method Illumina’s popular 450K human methylation
to whole genome sequencing data from 21 array. Unsurprisingly, the majority (>95%) of the
breast cancer samples. Eighteen putative significant differentially variable CpG sites in all
cis-regulatory mutations were identified cancer datasets tested showed more variability
in a number of samples that significantly in the cancer samples compared to the normal
increased the expression of the mapped gene samples. Interestingly, in the aging dataset,
when compared with other samples without 97% of the significant differentially variable
the mutation.This tool will be invaluable to CpGs were more variable in centenarians than
researchers seeking to prioritise large lists in newborns. A closer look at the genomic
of mutations for evaluating their functional composition of the significant CpGs revealed
relevance. that a greater proportion of the top ranked
differentially variable CpGs were found in
islands compared to differentially methylated
Poster 18 CpGs, a phenomenon that was consistently
observed across all datasets. DiffVar is available
in the Bioconductor R package missMethyl.
Dr Belinda Phipson, Dr Alicia Oshlack
DiffVar: A new method to test for

differential variability applied to DNA Poster 19
methylation in cancer and aging
Abstract Dr Ashfaqur Rahman, Dr Andrew Hellicar, Dr

DNA methylation is crucial for normal Daniel Smith, Dr John Henshall
development, however methylation changes
are known to accumulate with age; and Allele Frequency Calibration in Pooled DNA
samples: A Machine Learning Approach
aberrant methylation patterning is associated
with many diseases. During both tumour Abstract
development and aging, a global loss of We present a machine learning method
genome wide DNA methylation combined with to estimate allele frequency for SNP based
gains in CpG island promoter methylation has genotyping of DNA pools. DNA pooling is a
been observed. To date, the main focus when method commonly used to reduce the costs of
analysing DNA methylation data has been on genome wide association studies. DNA samples of
detecting differences in the mean levels of multiple individuals are combined into pools and
methylation between sample groups. Recently, genotyped. The raw array output is directly used
it has been speculated that identifying features to compute allele frequency of each SNP.
that differ in terms of variability may be just as
The raw array output from pooled DNA samples
relevant for understanding disease phenotypes.
is subject to greater genotyping inaccuracies
Methods for detecting differential variability
than individual samples and it is required to
as opposed to differential methylation have
adjust the allele frequency estimates. We present
not been well addressed. Our new method,
a supervised machine learning approach to
DiffVar, is inspired by Levene’s test and based
translate raw array output to allele frequency.
on an empirical Bayes hierarchical framework.
Given the genotypes of individuals that constitute
DiffVar is computationally efficient while being
the pool, we compute the true allele frequency
robust against non-normality and outliers.
and train a machine learning method to produce
The linear modelling framework allows any
a mapping between raw array outputs to the
experimental design to be accommodated.
true allele frequency. We then use this map to
75
Handbook
2014 AGTA
Conference
genotype pools with no prior of the individuals AGP is crucial for determining clinical strategies
constituting the pools. for intervention and prevention in at-risk
Two experiments were conducted each individuals. Analysis of a cytokine panel in a
using the same Sequenom iPLEX panel which longitudinal Finnish cohort (N = 2200) showed
generated results for 61 SNPs. The first larger an association between low-grade chronic
experiment genotyped 1041 individuals along inflammation with (future) elevated AGP. Next,
with 22 pools samples. The pool sizes varied we identified AGP-associated transcriptional
between 18 to 26 individuals all drawn from networks through weighted gene coexpression
the 1041 individuals. A second experiment network analysis on whole blood samples.
genotyped a sub set of 78 individuals from the Four transcriptional networks were found to
first experiment. We compared the performance be significantly associated with AGP, and
of the proposed calibration method with some of these sub-networks and their associations
the existing results. The RMSE error obtained using replicated in an independent cohort.
no calibration, commonly used piecewise linear Adjustment for lipid associated variation
transformation, and the proposed local–global identified one sub-network, enriched for host
learner fusion method are 0.135, 0.120, and response to bacterial and fungal infection.
0.080 respectively. Further evaluation revealed that its core genes
code for products released during neutrophil
degranulation, and Mendelian Randomisation
Poster 20 was used to infer causal or reactive
relationships between this sub-network and
AGP. Our study provides a general framework
Mr Scott Ritchie, Dr Peter Würtz , Ms Artika Nath, for integrative analysis of multi-omic data
Dr Gad Abraham, Dr Antti Kangas , Dr Aki sources, identifies molecular targets for follow-
Havulinna, Dr Pasi Soininen, Dr Kristiina Aalto, Dr up studies, and suggests prolonged innate
Johannes Kettunnen, Dr Michael Inouye immune response may be a driver of increased
mortality risk.
Integrative omics analysis elucidates
role for inflammation and neutrophil
degranulation in glycoprotein-associated
all-cause mortality risk
Abstract
Integration of molecular information captured
from high-throughput platforms has shown
immense promise for the elucidation of novel
biological processes underlying complex
disease phenotypes. Here, we integrate
genomic, transcriptomic, metabolomic,
and cytokine data in two Finnish population
cohorts to identify biological processes
underlying elevated circulating alpha-1-
acid glycoprotein (AGP) levels, a recently
discovered biomarker for 5-year all-cause
mortality risk. Circulating AGP is influenced by
a range of physiological effects; for example
AGP increases in response to inflammation,
infection, or tissue-injury, and interacts with a
variety of drugs. Identifying and understanding
the precise biological processes underlying
76
Handbook
2014 AGTA
Conference
Poster 21 Poster 22
Matthew E. Ritchie, Ruijie Liu, Gordon K. Smyth Alan F. Rubin, Lea M. Starita, Stanley Fields,
Terence P. Speed, and Douglas M. Fowler
Why weight? Variance modelling for
designed RNA-seq experiments Statistical methods and software for
functionally characterizing every single
Abstract mutation in Your Favorite Gene
Outlier samples are relatively common in
Abstract
RNA-seq experiments and the root cause
of such variation is generally unknown. In The use of high-throughput DNA sequencing
small experiments, the analyst is left with the has rapidly expanded catalogues of normal
difficult decision of what to do: removing the and disease-associated variation, but the
offending sample may reduce variation, but functional consequences of most mutations
at a cost of reducing power, which can limit are unknown. In deep mutational scanning,
our ability to detect biologically meaningful selection for protein function applied to
changes. A compromise is to use all of the a library of protein variants is combined
available data, but to down-weight the with high-throughput DNA sequencing to
observations from the outlier sample in the directly measure the activity of hundreds of
analysis. In this presentation we describe thousands of variants of the protein easily
a statistical approach that allows this by and cheaply. This approach helps to bridge
modelling heterogeneity at both the sample the gap between variant identification
and observational level in the differential and interpretation, allowing researchers to
expression analysis. Using both simulations and elucidate sequence-function relationships
real data, we tease apart scenarios where this at high resolution. The resulting data can
strategy leads to a more powerful analysis from be used in a variety of contexts, from aiding
those when it is better to discard the outlier the assessment of clinical variants to guiding
altogether. Our approach is implemented in protein engineering. Despite the growing
the open-source limma package available popularity of deep mutational scanning,
from Bioconductor (http://www.bioconductor. there are few tools and no formal statistical
org). methods available to help analyze these
complex datasets. Here we present a novel
method for assigning functional scores and
statistical significance to all variants in a
deep mutational scanning dataset based
on weighted regression. Our methods are
implemented as part of Enrich 2, a software
package that makes the initial data analysis
accessible to experimental biologists while
providing an extensible framework for
bioinformaticians manipulating these large
datasets. We share our results from applying
these methods to deep mutational scans of
BRCA1, the WW protein-binding domain of
YAP65, and other targets. Enrich 2 fully supports
barcode sequencing, complex experimental
designs involving controls and biological
replicates, and deep mutational scans of
noncoding sequences, such as structural RNAs
or regulatory elements.
77
Handbook
2014 AGTA
Conference
Poster 23 Poster 24
Mr Rick Tankard, Dr Carol Dobson-Stone, Dr Paul Wang, Dr Wendy Parker, Dr David

Professor Martin Delatycki, Associate Professor Yeung, Professor Susan Branford, Dr Andreas
David J Amor , Dr Katherine B Howell, Professor Schreiber
Elsdon Storey, Dr Richard J Leventer, Associate
Optimising multi-caller approaches to
Professor John Kwok, Associate Professor Paul J
somatic variant calling
Lockhart, Associate Professor Melanie Bahlo
Abstract
A statistical method to detect pathogenic
Cancer research has benefited greatly from
short-tandem repeat expansions in
Massively-Parallel Sequencing recent improvements in next-generation
sequencing methods that have led to
Abstract significantly increased sequencing depth at
Expansions of short-tandem repeats (STR) relatively low cost. In particular, detection
in humans are responsible for over twenty of somatic variants, which often require
neurological Mendelian disorders including matched normal-tumour samples and greater
spinocerebellar ataxias, intellectual disabilities, detection sensitivity (due to low tumour allele
epilepsies and Huntington’s disease. Current fraction) than germline variants, has become
methods of analysing massively parallel technically and economically feasible.
sequencing (MPS) for STR size are focused
Several dedicated somatic variant-calling
on genotyping alleles that are smaller than
algorithms have been developed recently.
read-fragment size, whereas pathogenic STR
Studies that examine these variant callers
expansions can be larger than the fragment
show that there is a much greater discrepancy
size, thus methods are needed to detect
between somatic variant callers than
pathogenic sized STRs.
germline variant callers. In most cases, the
We are building a statistical classifier to recommended approach is to use multiple
determine, in an individual with a potential callers and rely on their overlap to gauge
repeat expansion disorder sequenced with confidence for variant calling.
MPS, which STR loci have expansions larger
Using data from tumour and normal samples of
than those observed in a healthy population.
chronic myeloid leukaemia patients, we found
Here we explore a chi-squared statistic based
that while variants with high caller consensus
on binning of read alignments over STR loci.
are much more likely to be true variants (in
As the binning is likely to break independence
agreement with other studies), there exist many
assumptions required for a Pearson’s chi-
validated variants that are likely biologically
squared test, we perform permutation testing
relevant but have low caller consensus.
to derive the distribution of the chi-squared
statistic and generate meaningful p-values. As With the aim to achieve better understanding
MPS data of samples with known pathogenic of the differences between caller behaviour,
expansions is difficult to obtain, we instead we carried out in-depth analysis of eight
assess the performance with 400 microsatellite currently available somatic variant callers
markers (ABI PRISM Linkage Mapping Set v2.5) (MuTect, SomaticSniper, Seurat, Shimmer,
of thirty-two samples, four of which have EBCall, Strelka, Virmid, and VarScan), and also
Illumina whole-genome sequencing (WGS). The examined the impact of various processing
largest microsatellite marker allele across the steps, such as duplicate read removal and
thirty-two samples serve as ‘expansions’, with indel realignment. We will report these
those largest in an MPS sample used to assess findings and also make recommendations on
the power and type I error of our classifier to methods that can improve the accuracy and
detect smaller expansions. confidence of somatic variant calling.
78
Handbook
2014 AGTA
Conference
divided equally among all overlapping exons.

Poster 25 Here we investigate how a DEXseq analysis is
affected by modifying the count summerisation
process to allow only one count per read
Dr Katrina Bell, Dr Alicia Oshlack
compared to multiple exon counts per read.
Differential exon usage detection with RNA-
seq using altered count summarisation
options Poster 26
Abstract
RNA-seq is an invaluable tool for detecting
Leah Alcock, Scott Kuersten, Jim Pease, Agata
both differential gene expression and
Czyz, Anupama Khanna, Dixie Gabel, Victor
alternative isoform use between experimental
Ruotti, Fraz, Syed, Ramesh Vaidyanathan
conditions, tissues or cells. Alternative splicing
is a powerful form of gene regulation that Whole-Genome Bisulfite Sequencing (WGBS)
has received less attention than gene An improved “post-bisulfite” conversion
differential expression in the past. Currently library construction method from low
gDNA inputs
there are two main approaches for detecting
differential isoform expression using RNA- Abstract
Seq data. One requires the reconstruction of Epicentres continued efforts to innovate, has
each transcript with subsequent testing for led to the development of a novel library
differential expression between samples (e.g. construction method called EpiGnome™
Cufflinks followed by CuffDiff). The second Methyl-Seq, for preparing sequencing libraries
method utilizes a read-count approach at the from bisulfite-converted genomic DNA.
exon level in order to determine genes with Current whole-genome bisulfite sequencing
differential exon usage between samples (e.g. methods require substantial amounts of
Bioconductor; DEXseq, EdgeR-spliceVariants starting DNA to compensate for the loss due
and Voom -diffSplice). to bisulfite-mediated DNA degradation.
The differential exon usage (DEU) approach Additionally, conventional methods, involve
taken by the DEXseq method has several laborious DNA shearing, ligation of methylated
advantages. Firstly, the difficult task of sequencing adaptors and bisulfite conversion
reconstructing transcripts and estimating of unmethylated cytidine nucleotides. The
expression levels is not required for testing. main challenge is that approx. 90% of the
Secondly, the count based approach in adaptor tagged DNA is degraded during the
DEXseq builds on well-established rigorous bisulfite conversion step. As a consequence,
statistical testing techniques developed a majority of DNA fragments contained in
for RNA-seq analysis. However, in order for the adaptor tagged sequence library lose at
testing for differential DEU, DEXseq requires least one flanking adaptor rendering them
the RNA-seq reads to be summarised into a incapable of being sequenced. To address
count matrix containing the number of reads and overcome these issues, the EpiGnome™
overlapping known exons. An RNA-seq read Methyl-Seq kit “post-bisulfite” method features
has the potential to overlap multiple exons and bisulfite conversion of cytosines prior to
as sequencing read length is increasing, this addition of adaptor sequences required for
is becoming a more common occurrence. It cluster generation and sequencing. Thus DNA
is not clear however how these reads should fragments, independent of size and position
be treated. One approach is to assign a of bisulfite mediated strand breaks, remain
single read to each overlapping exon and as viable templates for library construction
count it multiple times. Other options include and sequencing. This novel approach exhibits
only counting the read in the first exon or increased sensitivity and efficiency, along with
proportional counting where a read is only improved coverage for detecting methyl-
counted once per experiment, with the count cytidine nucleotides.
79
Handbook
2014 AGTA
Conference
differentially-methylated probes (314) showing a

Poster 27 higher level of DNA methylation in term infant’s
blood compared to both preterm infants’ and
adult blood, suggesting that developmentally-
Associate Professor Jeffrey Craig, Dr Mark
associated epigenetic trajectories may not
Cruickshank, Dr Christiane Theda, Professor
always be linear.
Peter Davis 3, Dr Penny Sheehan, Dr David
Martino, Dr Alicia Oshlack, Ms Yun Dai, Dr We present evidence for widespread
Richard Saffery , Professor Lex Doyle methylation differences between extreme
preterm and term infants at birth that are
Epigenetic changes in survivors of preterm largely resolved by 18 years of age. These
birth: effect of gestational age and results are consistent with methylation changes
evidence for a long term legacy
associated with blood cell development,
Abstract immune induction, cellular composition and
Preterm birth confers a high risk of adverse age at these time points. Finally, using this small
long-term health outcomes for survivors, yet the sample size we identified a number of sites
underlying molecular mechanisms are unclear. that were persistently altered in the preterm
We hypothesized that effects of preterm group and may be associated with a long term
birth can be mediated through measurable epigenetic legacy of preterm birth.
epigenetic changes throughout development.
We therefore used a longitudinal birth cohort
to measure the epigenetic mark of DNA Poster 28
methylation at birth and 18 years comparing
survivors of extremely preterm birth with infants
born at term. Dr Christoffer Flensburg, Dr Alicia Oshlack
Using 12 extreme preterm birth cases and 12 A comparison of control samples for ChIP-
matched, term controls, we extracted DNA seq of histone modifications
from archived neonatal blood spots and blood
Abstract
collected in a similar way at 18 years of age.
The advent of high-throughput sequencing
DNA methylation was measured at 347,789
has allowed genome wide profiling of histone
autosomal locations throughout the genome
modifications by Chromatin ImmunoPrecipitation
using Infinium HM450 arrays. Representative
(ChIP) followed by sequencing (ChIP-seq). In
methylation differences were confirmed by
this assay the histone mark of interest is enriched
Sequenom MassArray EpiTyping.
through a chromatin pull-down assay using
At birth we found 1,555 sites with significant an antibody for the mark. Due to imperfect
differences in methylation between term and antibodies and other factors, many of the
preterm babies. At 18 years of age, these sequenced fragments do not originate from
differences had largely resolved, suggesting the histone mark of interest, and are referred
that DNA methylation differences at birth are to as background reads. Background reads
mainly driven by factors relating to gestational are not uniformly distributed and therefore
age, such as cell maturity and to some extent, control samples are usually used to estimate the
cell composition. Using matched longitudinal background distribution at any given genomic
samples, we found suggestive evidence for an position. The Encyclopedia of DNA Elements
epigenetic legacy of preterm birth, identifying (ENCODE) Consortium guidelines suggest
persistent methylation differences at several sequencing a whole cell extract (WCE, or “input”)
genomic loci. Longitudinal comparisons sample, or a mock ChIP reaction such as an IgG
uncovered a significant overlap between sites control, as a background sample. However, for
that were differentially-methylated at birth and a histone modification ChIP-seq investigation it is
those that changed with age. We found an also possible to use a Histone H3 (H3) pull-down to
unexpectedly high number of birth- and age- map the underlying distribution of histones.
80
Handbook
2014 AGTA
Conference
Here we generated data from a

hematopoietic stem and progenitor cell Poster 30
population isolated from mouse foetal liver
to compare WCE and H3 ChIP-seq as control
Miss Siobhan Hughes
samples. The quality of the control samples is
estimated by a comparison to pull-downs of Genomewide Analysis of Proteins that Bind
histone modifications and to expression data. to DNA and Regulate Gene Expression
We find minor differences between WCE and
Abstract
H3 ChIP-seq, such as coverage in mitochondria
Regulation of gene expression is a
and behaviour close to transcription start sites.
complicated process, subject to different
Where the two controls differ, the H3 pull-down
mechanisms operating at different levels.
is generally more similar to the ChIP-seq of
At the genomewide level, chromatin
histone modifications. However, the differences
immunoprecipitation and next-generation
between H3 and WCE have a negligible
sequencing are being used to interrogate
impact on the quality of a standard analysis.
the coincident and allele-specific binding of
the proteins CTCF, Cohesin, ATRX and MeCP2
in the murine brain, a tissue that expresses a
Poster 29
significant proportion of imprinted transcripts.
Identifying regions co-binding these proteins
Dr Aliaksei Z. Holik, Laura A. Galvis Vargas, Julie will help generate a model to understand
Pasquet, Aaron T. L. Lun , Dr Marnie E. Blewit, how these proteins influence gene expression,
Dr Matthew E. Ritchie, Dr Marie-Liesse Asselin- particularly at imprinted loci. Using these tools
Labat we aim to understand more about the proteins
that participate in the genomic landscape
Ezh2 is essential for lung development, around imprinted loci and drive this unusual
maintenance of epithelial cell identity and
mode of gene regulation. These loci will act as
cell fate determination
models for studying DNA binding proteins and
Abstract their roles in transcription.
The role of epigenetic regulation in At the level of the individual locus, mechanisms
developmental processes has become of gene regulation are being investigated using
increasingly evident. The polycomb repressive imprinted retrogenes as a model. Retrogenes
complex 2 (PRC2) has been shown to play are transcriptionally active intronless genes
a fundamental role in controlling stem cell located within an intron of a ‘host’ gene. The
function in a number of organs. However, its role of epigenetic factors influencing gene
involvement in lung development, as well as transcription are being investigated at the
the role of epigenetic regulation in general, Mcts2/H13 locus. Expression of an intronic
remains poorly understood. We explored the retrogene can cause premature termination
role of the catalytic subunit of PRC2, Ezh2, of a ‘host’ transcript from the same allele. Our
during lung morphogenesis. Loss of Ezh2 in the hypothesis is that this premature termination
lung epithelium led to defective lung formation is caused by transcription of the retrogene
and perinatal mortality. We show that Ezh2 interfering with host gene transcription. We
is critical for airway lineage specification have designed a construct based on this locus,
and alveolarization. We use gene expression which will allow us to regulate the expression
profiling of embryonic lung mesenchyme and through the retrogene/internal promoter to
epithelium deficient in Ezh2, combined with study this in more detail. These studies will
ChIP-seq analysis of H3K27 tri-methylation, as provide a mechanistic component to our
well as publicly available gene expression and whole genome analyses.
histone modification datasets to elucidate
potential mechanisms of Ezh2 function in
embryonic lung development.
81
Handbook
2014 AGTA
Conference
We are currently building on this research to

Poster 31 develop methylation biomarkers that can be
used to estimate the proportion of nTreg in
human blood, which may have diagnostic
Dr Jovana Maksimovic, Dr Yuxia Zhang, Mr
and prognostic applications for autoimmune
Gaetano Naselli, Junyan Qian, Dr Michael
disease.
Chopin, Dr Marnie E. Blewitt , Prof. Leonard C.
Harrison, Dr Alicia Oshlack
Epigenetics, T-cells and autoimmunity: A Poster 32

bioinformatics story
Abstract Michelle French, Anne O’Connell, Sara Edwards,

Autoimmune diseases result in an inappropriate Avijit Haldar, Phil Farquhar, Ken Dodds, Sue
immune response against healthy cells. They Galloway, Rudiger Brauning, Peter Johnstone,
are likely caused by a combination of genetic George Davis, Jenny Juengel
predisposition and environmental triggers.
Epigenetic phenomena are thought to be Mutations in LEPR gene are associated with
reproductive traits in Davisdale sheep
the link between the environment, genetic
background and development of autoimmune Abstract
disease. Whole genome sequencing of four key sires
Regulatory T cells (Treg) are required to prevent from the highly prolific Davisdale sheep line
the emergence of autoimmune disease and were used to identify mutations in candidate
evidence suggests that epigenetic regulation is genes known to be involved in the regulation of
required to establish their ultimate phenotype reproductive performance. This identified three
and function. To explore the methylation single nucleotide polymorphisms (SNPs) that
landscape of Treg we analysed genome-wide change the amino acid sequence of the Leptin
methylation in human naïve Treg (rTreg) and Receptor (LEPR) gene. A sequenom assay was
conventional naïve CD4+ T cells (Naïve) using designed and we genotyped 750 ewes with
Illumina 450k arrays. By taking into account the records for ovulation rate, first service conception
differences between individuals and performing rate and partial failure of multiple ovulations;
both probe-wise and regional differential 600 of these ewes also had records for puberty
methylation analysis we detected 2,315 age. The aim of this study was to determine if
differentially methylated CpGs between these the SNPs in the LEPR gene were associated with
two cell types, many of which clustered into reproductive traits. Ewes homozygous for the LEPR
127 regions of differential methylation (RDMs). gene mutations either failed to undergo puberty
Further analysis showed that the RDMs were during their first year, or those that did attain
enriched for putative FOXP3 binding motifs. puberty were 18 days older than ewes which did
Bioinformatic analysis of publicly available not have the LEPR mutations. Age at puberty
data also revealed that CpGs within known was intermediate between the two homozygous
FOXP3-binding regions were hypomethylated groups in ewes heterozygous for the mutations.
and that there was an inverse relationship Ewes that were homozygous for mutations in the
between promoter RDM methylation and gene LEPR gene had a 15% decrease in ovulation rate
expression. Thus, hypothesising that methylation (P<0.001), and a 12% decrease in their first service
limits access of FOXP3 to its DNA targets we conception rate (P<0.01). Ewes that ovulated
proceeded to show that increased expression three ova had increased partial failure of multiple
of the immune suppressive receptor TIGIT was ovulations (P<0.01), and had 0.2 fewer lambs at
associated with hypomethylation and FOXP3 mid-pregnancy and at birth (P<0.01) compared
binding at the TIGIT locus. to ewes that are wild type. Our studies have
shown that poorer reproductive performance
in the Davisdale line is strongly associated with
homozygosity for mutations in LEPR.
82
Handbook
2014 AGTA
Conference
Poster 33 Poster 34
Mr Amali Thrimawithana, Mr Bernd Steinwender, Dr Christy Vander Jagt, Dr Amanda

Mr Ross Crowhurst, Mr Richard Newcomb Chamberlain, Professor Mike Goddard, Dr Leah
Marett, Ms Lakshmi Krishnan, Professor Ben
Exploring the pheromone glands of native
Hayes
New Zealand leafroller moths using de
novo transcripts preliminary insights A Bovine Gene Expression Atlas
Abstract Abstract
Many of the enzymes involved in pheromone The subset of expressed genes within a cell
biosynthesis pathways in moths are yet to be convey the unique properties of different
discovered. We have used a transcriptomics- cell- and tissue-types. In the cow, the
based approach to help to gain insights into structure of this ‘expression space’ is largely
the pheromone biosynthesis pathway, with unknown. Next-generation sequencing allows
a longer-term goal of understanding the comprehensive transcriptome surveys to be
involvement of these genes within this pathway performed, providing insight into tissue-specific
in chemical signalling. We explored the gene expression. In this study we employed
pheromone glands of New Zealand leafroller RNA-sequencing (RNA-seq) to construct a
moths. New Zealand leafroller moths include bovine gene expression atlas, generating
species from the genera Ctenopseustis and a snapshot of gene expression at a single
Planotortrix. We have focused our research time point across 18 different tissue types in a
on four species (C. obliquana, C. herana, P. lactating dairy cow.
octo and P. excessana). These species pose
100 base paired end RNA-seq reads were
a significant threat to important horticultural
generated on an Illumina HiSeq2000 from each
crops in New Zealand because of their highly
tissue (in triplicate). Tophat2 was used to map
polyphagous nature. The sibling species are
reads to the Ensembl annotation of UMD3.1
similar in appearance, but are known to be
bovine genome assembly. On average 92%
reproductively isolated in the wild because
of reads aligned per sample (>70% uniquely
of distinct sex pheromones. Therefore
mapped). Differential expression analysis using
exploration of the pheromone biosynthesis
DESeq and EdgeR identified tissue-specific
pathways in these species will be beneficial
transcriptional trends and demonstrated that
in understanding the differences among
of the 18 tissues, the two brain tissues had the
these species. To achieve this we sampled
greatest number of significantly differentially
the transcriptomes of isolated pheromone
expressed genes (when compared to average
glands from each species and have de novo
gene expression). The relationship between
assembled transcripts using a multi-assembler
tissues was investigated via hierarchical
approach. We highlight and discuss the
clustering. Tissues clustered together reflecting
preliminary insights we gained from this multi-
their biological relationship (i.e. all muscle tissues
assembler approach.
clustered together), providing some validation
of results. Functional annotation of differentially
expressed genes further validated results, with
many biological processes and pathways
identified as being significantly enriched
(p<0.01) already having an established role
in those particular tissues (i.e. the top most
significant Gene Ontology term associated with
differentially expressed genes in both white skin
and black skin was ‘epidermis development’).
83
Handbook
2014 AGTA
Conference
The bovine gene expression atlas generated parasitic plants. In addition to advancing
by this study will advance our understanding our understanding stress in woody hosts it will
of the dynamics and complexity of bovine increase our understanding of the link between
transcription and provide a valuable resource belowground and aboveground events and
for future research on functional genomics in provide genetic information that can be used
the bovine species. for studying natural communities.
Poster 35 Poster 36
Mr Bjorn Dueholm, Dr Bjorn Hamberger, Chol-Hee Jung, Professor Mohan Singh,

Professor Philip Weinstein, Dr Susan Semple, Professor Prem Bhalla
Professor Birger Lindberg Moller
Dynamic activities of novel transcripts in
Comparative transcriptomics of Acacia leaf and shoot apical meristem of soybean
acuminata parasitised vs. non-parasitised during floral transition
by the hemiparasitic shrub Santalum
spicatum Abstract
The analysis of high-throughput sequencing
Abstract of RNAs is capable of revealing the
Acacia acuminata (Raspberry jam tree) details of transcriptional landscapes in an
is common to Western Australia and often unprecedented level. We recently conducted
found parasitised by the hemiparasitic shrub whole genome transcriptome analysis of
Santalum spicatum (Australian sandalwood). soybean, one of the major crop plants, to
Australian sandalwood takes up water, uncover the dynamic transcription and
nutrients, and carbon from other plants regulation during the early floral initiation
through specialised organs on the roots called process (Wong et al., 2013, PLoS ONE) focusing
haustoria. This parasitism puts the acacia host on protein coding genes. Soybean is a
under some level of biotic stress. Studies have paleopolyploid that has been subjected to at
shown a variety of different physiological least two rounds of whole genome duplication
stress-responses from the plant parasitism in the events. We are interested to investigate
acacia host; however, how the host responds involvement of alternative splicing, antisense
on a transcriptional level has not yet been transcripts and intergenic transcripts during
investigated. floral initiation in soybean after inductive
RNA-Seq is a powerful method for the short-day treatment. We identified ~107,000
investigation of gene expression in non-model transcripts from ~56,000 loci in the soybean
organisms and can be used for comparative genome using a total of ~120 million paired-
transcriptomics between two or more types of end sequences from leave and shoot apical
treatments. Experimental plantations in which meristem samples taken at different time-points
Raspberry jam tree grows under the same after short-day treatment (floral-inducing). We
abiotic conditions either alone or together observed that the shoot apical meristem utilizes
with Australian sandalwood will be utilized a greater repertoire of splicing variants than
for investigating the differential expression leaf and that the use of alternative splicing
of genes in the acacia host under the two isoforms changes dynamically depending on
treatments. developmental stages and/or tissue types.
Multiple-isoform loci are enriched for antisense
Parasitic plants have been found not only to
transcripts compared to single-isoform loci.
affect the hosts but also indirectly to affect
The same sense and antisense transcript
other organisms associated with the hosts.
pair appears to have a different relationship
This study is the first to utilize RNA-Seq for
in different tissues, and the same antisense
studying stress-induction in woody hosts by
84
Handbook
2014 AGTA
Conference
transcript may interact differently with different Now, with the cost per marker continuing to
sense transcripts in the same locus. Overall, our drop, we investigate the feasibility of genome-
study improves the annotation of the soybean wide association and genomic selection for
genome by identifying numerous novel improving essential oil yield in E. polybractea.
transcripts and demonstrates the complex We compare traditional breeding and marker-
regulation of splicing variants with or without assisted selection for essential oil yield, and how
an association with antisense transcripts during the use of whole genome re-sequencing and
floral initiation. genomic selection may further improve selection
of elite individuals.
Poster 37
Poster 38
David Kainer, Dr Amanda Padovan, Hamish

Webb, Professor William Foley, Dr Carsten Dr Jafar Sheikh Jabbari, Dr Rust Turakulov, Dr
Kulheim Kirby Siemering, Mr Matthew Tinning, Ms Maria
Pazos-Navarro, Dr Daniel Real, Dr Matthew N.
Towards Genomic Selection for essential
Nelson
oil yield in Eucalyptus and Melaleuca
Efficient Genotyping-by-Sequencing Enables
Abstract
Construction of the First Linkage Map for
The yield of essential oil in commercially the Non-Model Legume Forage Species,
harvested Myrtaceae species (e.g. Eucalyptus Bituminaria bituminosa var. albomarginata
polybractea, Eucalyptus loxophleba (tedera)
and Melaleuca alternifolia) is a complex
Abstract
quantitative trait which is itself composed of
Single-nucleotide-polymorphisms (SNPs) identified
multiple quantitative traits such as foliar oil
using genotyping-by-sequencing (GBS) methods
concentration, biomass and adaptability.
are increasingly used in many fields of genetics
These traits often show large natural variation
and population genomics. These methods differ
and some are highly heritable. Foliar essential
in nucleotide diversity during initial sequencing
oils are mostly comprised of mixes of mono-
cycles that affects quality and quantity of reads,
and sesquiterpene compounds and so
number of sampled digestion fragments (tags)
recent association studies have investigated
and sequencing depth required for confident
well characterised genes of the terpene
genotype calls. Here we introduce an improved
biosynthetic pathway. Analysis of transcript
GBS method based on size-selection of double-
abundance and allelic diversity has revealed
digested adapter-ligated DNA which increases
that essential oil yield is likely to be controlled
sequencing yield and quality by producing high
by large numbers of quantitative trait loci
diversity libraries. We applied this technique for
(QTLs) that range from a few of medium/large
genotyping a subset of an F2 population from
effect to many of small effect, with much of
Bituminaria bituminosa var. albomarginata
the variation occurring beyond the major
(tedera), an emerging drought-tolerant forage
genes of the biosynthetic pathway.
legume species. By sequencing a barcoded
Genome-wide association studies (GWAS) library prepared from 45 progeny and parents
are often used to investigate the genetic in one lane of HiSeq we obtained over 150 M SE
architecture of complex quantitative traits reads which were analysed using Stacks pipeline
across the whole genome to explain as much resulting in 845 high-quality SNP markers scored
variation as possible. However, in naturally in all samples. Linkage mapping identified 11
outcrossing forest trees the feasibility of GWAS linkage groups, close to the haploid chromosome
has been limited by the requirement for very number of this species (n=10). Analysis of synteny
high density genotyped markers due to the via BLASTn between the genomes of tedera
short range of linkage disequilibrium (~100 bp). and its close relative soybean using tedera GBS
85
Handbook
2014 AGTA
Conference
sequence tags revealed the presence of 264

(31.2%) GBS marker tags with one or more Poster 40
significant (<1e-10) matches in the soybean
genome and that marker order was conserved
Dr Jeremy Shearman, Ms Duangjai Sangsrakru,
in large chromosomal blocks between the two
Ms Panthita Ruang-areerate, Ms Chutima
genomes. This finding, along with application
Sonthirod, Ms Pichahpuk Uthaipaisanwong, Ms
of the method to the remaining progeny, will
Thippawan Yoocha, Ms Supannee Poopear,
facilitate the transfer of genomic information
Dr Kanikar Theerawattanasuk, Dr Somvong
between the well-resourced model genome of
Tragoonrung, Dr Sithichoke Tangphatsornruang
soybean and the ‘orphan’ genome of tedera.
Assembly and analysis of a male sterile
rubber tree mitochondrial genome reveals
Poster 39 DNA rearrangement events and a novel
transcript
Abstract
Mnasri Sameh, Saddoud Dabbebi Olfa, Ben
The rubber tree, Hevea brasiliensis, is an
Saleh Mohamed, Professeur Ferchichi Ali
important plant species that is commercially
Microsatellite marker-based identification grown to produce latex rubber in many
and genetic relationships of millennium countries. The rubber tree variety BPM 24
olive cultivars in Tunisia exhibits cytoplasmic male sterility, inherited
from the variety GT 1. We constructed the
Abstract
rubber tree mitochondrial genome of a
Microsatellite markers were used to characterize
cytoplasmic male sterile variety, BPM 24, using
the millennium olive cultivars localized in
454 sequencing, including 8 kb paired-end
nine different archeological sites in Tunisia.
libraries, plus Illumina paired-end sequencing.
Thirty genotypes were considered for genetic
We annotated this mitochondrial genome
fingerprinting using 10 pairs of microsatellite
with the aid of Illumina RNA-seq data and
primers. The number of alleles per locus ranged
performed comparative analysis. We then
from 3 to 5, with a mean of 3.7 alleles per
compared the sequence of BPM 24 to the
primer pair (a total of 37 alleles). The observed
contigs of the published rubber tree, variety
heterozygosity ranged from 0.4 to 1, while the
RRIM 600, and identified a rearrangement
expected heterozygosity varied between 0.37
that is unique to BPM 24 resulting in a novel
and 0.74. The polymorphism information content
transcript containing a portion of atp9. The
values ranged also from 0.37 to 0.74. The mean
novel transcript is consistent with changes
polymorphism information content value of 0.61
that cause cytoplasmic male sterility
for the SSR loci provided sufficient discriminating
through a slight reduction to ATP production
ability to evaluate the genetic diversity among
efficiency. The exhaustive nature of the search
the millennium cultivars. The UPGMA cluster
rules out alternative causes and supports
analyses using Jaccard’s index permitted a
previous findings of novel transcripts causing
segregation of the thirty millennium cultivars in
cytoplasmic male sterility.
three main groups and revealed that most of
the millennium cultivars grouped according
morphological parameters of the fruit and the
endocarp and no clear clustering trends were
observed according to their geographic origin.
As a sequel to the present work, new surveys
should be made in the archeological sites
localized in North and the Center of Tunisia to
sample more cultivars and to draw a clearer
picture of the diversity of the Tunisian millennium
olive germplasm.
86
Handbook
2014 AGTA
Conference
were restricted to TEs in the columella genome,

Poster 41 and the patterns or levels of DNA methylation
in expressed genes did not differ between the
cells types studied.
Mr Tim Stuart, Prof. Ryan Lister, Dr Taiji
Kawakatsu, Dr Natalie Breakfield, Dr Manuel As columella cells are terminally differentiated,
Valdes, Dr Joseph Nery, Dr Xinwei Han, Prof. mutations induced through TE activity are not
Robert Schmitz, Prof. Philip Benfey, Prof. Joseph inherited by any other cells, and consequently
Ecker would have a small impact on the overall
fitness of the plant. However, the enhancement
Single cell-type analysis reveals unique of TE silencing in the columella observed in this
patterns of DNA methylation in the root
study indicates that TE silencing is particularly
meristem
important in these cells. One possible
Abstract explanation for this seeming contradiction is
In the plant genome, methylation of cytosine that 24 nt smRNAs generated in the columella
bases (DNA methylation) plays a critical role are transported to the neighbouring root
in transcriptional silencing of transposable stem cell to reinforce transcriptional silencing
elements (TEs) to protect genome stability. of TEs through RdDM, similar to processes
Multiple pathways exist in Arabidopsis thaliana thought to occur between reproductive cells
for the establishment and maintenance of and their non-generative companion cells.
DNA methylation, including the RNA-directed This establishes the columella as potential
DNA methylation (RdDM) pathway utilizing a companion cells to the root stem cells.
24 nucleotide (nt) small RNA (smRNA) signal
to direct DNA methylation. Recently, single
cell-type analysis of Arabidopsis reproductive Poster 42
cells unveiled widespread changes in DNA
methylation patterns and smRNA production
in the reproductive companion cells, and Mr Anandhakumar Chandran, Mr Li Yue, Mr
that these changes play an important role in Seiichiro Kizaki, Ms Le Han, Dr Ganesh Pandian,
maintaining TE silencing in the gametes and Mr Syed Junetha, Mr Shinsuke Sato, Mr Junichi
developing embryo. Taniguchi, Dr Toshikazu Bando, Dr Hiroshi
Sugiyama
In plants, stem cell niches (meristems)
responsible for tissue growth and development High through-put Sequencing Guides the
exist in the root and shoot. To investigate Design of DNA Minor Groove Binding Small
whether a reconfiguration of DNA methylation Molecule with Improved Binding Specificity
marks similar to that of the reproductive cells Abstract
occurs in meristems, we used an integrated In chemical genomics and molecular biology,
analysis of DNA methylation, mRNA and smRNA a small molecule is an organic compound with
transcript abundance in five different cell types low molecular weight that is useful in regulating
of the root meristem. Our analysis revealed biological process. The development of small
that the columella, a group of terminally molecules that trigger targeted transcriptional
differentiated, short-lived cells situated directly activation could ensure better efficacy and
below the root stem cells, contains levels of a reduction of long-term side effects. Our
DNA methylation far higher than any other results show, an artificial small molecule named
Arabidopsis cell or tissue type reported to SAHA-PIP K containing the histone deacetylase
date. Hypermethylation of the columella inhibitor SAHA and the sequence-specific DNA
genome was coupled with an increase in minor groove binding hairpin pyrrole-imidazole
mRNA transcripts encoding components of the polyamides (PIPs), could rapidly stimulate
RdDM pathway, as well as elevated levels of 24 the meiotic process regulating PIWI pathway
nt smRNAs needed for RdDM. These changes
87
Handbook
2014 AGTA
Conference
marker genes in human dermal fibroblasts.

This first ever result on a small molecule Poster 43
that significantly induces the endogenous
regulation of the conserved germ cell genes in
Mr Arunkumar Padmanaban, Dr Eva Schmidt,
a human somatic cells opens up the panorama
Dr Adam Inche, Dr Ruediger Salowsky
of chemical mediated reprogramming of
somatic cell to a germ cell. But, whether SAHA- End-to-end sample quality control for next
PIP’s strong biological effects due to its binding generation sequencing library preparation
specificity in a broad context or else other is still and SureSelect Target Enrichment on the
Agilent 2200 TapeStation System
unclear. To characterize the binding specificity
of SAHA-PIPs, we carried out high-throughput Abstract
sequencing (Bind-n-seq) studies to define the Next generation sequencing (NGS) has
recognition motifs of PIP K (PIP without SAHA) been widely adapted to address intriguing
/ SAHA-PIP K -DNA binding in a genome sized biological questions and understand
sequence space. SAHA-PIP K showed, high mechanisms at the genome level. New
affinity binding compared with PIP K. This developments in sequencing methods and
induced us to redesign SAHA-PIP without SAHA high-throughput technologies has reduced
and with the moiety in the non-core binding the timeline for genome sequencing from
region of the ten-ring hairpin polyamide. The weeks to days. Even though the cost of
results showed the moiety can enforces the NGS has significantly reduced, the upfront
closed binding of PIP. sequencing library preparation is still complex
and time consuming. Several companies
offer library preparation workflows to help
reduce complexity. In addition, the targeted
enrichment of the end libraries for specific
genomic subsets can further drive down cost
by increasing sequencing coverage. However,
library preparation workflows still need to be
carefully controlled and monitored to ensure
downstream sequencing success. The Agilent
SureSelect Target Enrichment solution offers
a complete, validated workflow of sample
preparation for target specific sequencing with
a range of capture sets including all exons,
panels like clinical research exome. Here we
present data to demonstrate the Agilent 2200
TapeStation system as an ideal and reliable
quality control (QC) platform for sizing and
quantification analysis of the NGS libraries. The
Genomic DNA ScreenTape assay assess the
starting genomic DNA sample for quality and
quantity, replacing independent measurement
systems with a single test, thereby saving
sample and time. The sizing and quantification
of the intermediate and final NGS libraries
can be reliably carried out using the D1000
ScreenTape assay. Thus, the 2200 TapeStation
system allows analysis of the starting genomic
sample (DNA), through to final QC of the DNA
libraries prior to sequencing.
88
Handbook
2014 AGTA
Conference
at all. The accuracy of the index need to be

Poster 44 further validated.
Inaddition, the consistency rate of the
Dr Li Zhou, Daniel Catchpoole Fluidigm panel is 100% within the plate due to
its qualitative feature, but it is not optimal to
Evaluation of Fluidigm SNPtrace™ Panel, compare the results across the plates, whereas
with a comparison to DNA QUAL and due to its qPCR feature, DNAqual can compare
bioanalyzer fragment analysis, for DNA
the readings across runs. All these results will be
quality assurance within the Tumour Bank
program further validated and elucidated.
These results form the basis of a better
Abstract
biospecimen QA solution for the Australian
Translational research is the key component
biobanking community. Furthermore, we will
to apply basic science findings into practical
discuss the ongoing implementation of these
medicine. High quality human biological
new QA platforms into daily biobank practice
specimens are the foundation of accurate and
and how these will benefit translational research
reproducible research result and will maximize
involving genomics.
the productivity of translational research. To
assure that the best samples are provided to
translational research, quality assurance (QA)
Poster 45
programmes are an essential part of biobanks’
routine.
To identify the most feasible technology for Associate Professor Richard Allcock
regular biobank QAP daily practice, two
A Better Way - AmpliSeq Exome for Clinical
technologies promoted from use in biobanking
Research
QAP have been compared, the Fluidigm SNP
trace panel (Fluidigm, USA) and DNAqual Abstract
(Eurobio, France). 93 samples, including 80 DNA Many laboratories are actively engaged
samples from 4 donors and 13 DNA samples in implementing large-scale sequencing
isolated from bone marrow aspirates (BMA) in clinical and diagnostic settings. There is
smeared on slides, were examined using both significant debate about the best approach
technologies. BMA slides had been stored at (gene-specific panels, WES, WGS), each
the room temperature and exposed to air of which has benefits and disadvantages.
for 2-14 years. In addition, to create a range In addition to accuracy and utility, other
of possible conditions occurring in biobank practical considerations must be taken into
daily practice, DNA samples from donors were account such as cost, scalability, practicality of
treated with multiple freeze-thaw cycles (0-20), implementation and infrastructure requirements.
varied snap delay period (0-21 days), radiation, In many settings, WES may strike the appropriate
sonication, heat, UV and mixed contamination balance between these factors.Since its
(MC) respectively. DNA qualities have been development, WES has made incremental
verified using Bioanalyzer DNA 7500 Kit (Agilent, advances but for the most part is still based upon
USA). hybridisation to biotinylated probes. Recently
The results showed that the Fluidigm panel a completely new way to enrich exomes was
could detect 50% MC between samples developed based on massively parallel PCR
accurately. The sensitivity of the contamination (AmpliSeq). In combination with the Ion Torrent
detection need to be further tested. It also Chef and Proton sequencer, the process is rapid,
reported the degraded DNA qualitatively flexible, scalable in a variety of settings and
instead of quantitatively. By contrast, DNAqual highly reproducible. it is now possible to routinely
can provide the quantitative DNA quality index and economically sequence 100X+ coverage
although it is not designed for detecting MC exomes).
89
Handbook
2014 AGTA
Conference
Exomes enriched by amplification differ

significantly from those enriched by
Tuesday
hybridisation. First, amplification is more specific 14 October
than hybridisation and hence very high on-
target rates (>94%) are can be achieved.
Second, there is less need to sequence deep
Poster 46
into introns, with primers sited closer to the
exons of interest. Third, analytical approaches
need to be modified as a PCR-based approach Nigel C Bennett, Maryam Jessri, Andrew J
produces PCR duplicates (by definition). Finally, Dalley, Camile S Farah
the process is extremely rapid, allowing us to go
from DNA to final analysed result in as little as Exomic molecular signatures segregate
oral epithelial lesions based on disease
36 hours, which potentially allows application
severity
in many areas not traditionally amenable to
exome level analysis.I will provide an overview Abstract
of the latest developments using the AmpliSeq To determine if the accumulation and genetic
Exome RDY kit and its application to clinical location of deleterious mutations within oral
research problems in Western Australia. epithelial lesions correlates with increasing grades
of epithelial pathology, 53 archival formalin fixed
paraffin embedded (FFPE) oral epithelial biopsies
inclusive of 5 oral squamous cell carcinoma,
31 epithelial dysplasia, and 11 samples with no
epithelial abnormality were examined using
next generation sequencing. Target-enriched
exome libraries were constructed from 500ng-3µg
gDNA using the SureSelectXT target enrichment
system (Agilent Technologies, CA, USA) for
SOLiDTM 5500XL multiplexed sequencing (Life
TechnologiesTM, CA, USA). Sequenced exomes
were mapped to the hg19 reference human
genome and variants called using the GATK
software. The numbers of variants in each gene
were compared in different groups of samples
using Kruskal-Wallis and sPLS-DA. All p-values
were corrected for multiple comparisons using
Bonferroni’s correction. Per sample, a mean
of 170,906,808±52,255,566 unique reads were
generated from which a mean of 79±7.8%
were on target with a mean coverage of
137X. A mean of 23,335±2,554 variants were
called per sample, from which synonymous
variants (mean=11,339±1,258) were excluded.
When tested by Kruskal-Wallis, lesions could be
significantly discriminated based on the number
of exomic variants (non-synonymous, frameshift
and non-sense). Multivariate analysis separated
the lesions based on their severity using two
components and 10 variables. An increased
number of exomic variants in different oral
epithelial pathological states is supportive of the
90
Handbook
2014 AGTA
Conference
field cancerization theory. Oral epithelial biopsies tumour tissues, suggesting DPM accumulation is
can be separated based on their severity using a general feature of carcinogenesis and could
the number of mutations in their exome. be used as a novel cancer biomarker.
Poster 47 Poster 48
Ms Magdalena Budzinska, Dr Thomas Tu, Dr Wenhan Chen, Devika Ganesamoorthy and

Fabio Luciani, A Nicholas Shackel Lachlan Coin
Genomic evolution of hepatocytes Detecting tumour DNA in blood plasma

towards hepatocellular carcinoma using allelic imbalance from read
countdata
Abstract
Hepatocellular Carcinoma (HCC) is associated Abstract
with accumulation of number of genetic Tumour cells undergo apoptosis release small
abnormalities. The vast majority of those amount of cell-free circulating tumour DNA
events, called passenger mutations, are non- (ctDNA) into blood throughout all stages
recurrent and dispersed across the cancer of cancer. These ctDNA carry the somatic
genome. A previous study has shown that mutations which have accumulated in the
deleterious passenger mutations (DPM) occur tumour cells, including copy number alteration
more frequently in tumour tissue compared (sCNA) of whole chromosomes (aneuploidy)
to surrounding non-tumour tissue. Therefore, and chromosome arms. Aneuploidy is also a
we hypothesise that there is a stepwise mark of high grade tumour development and
accumulation of DPM in HCC progression. To poor survival prognosis. We hypothesised that
address this, we aimed to characterise DPM in next generation sequencing of cell free DNA
whole exome sequencing (WES) datasets from in plasma may enable us to both detect the
patients in various stages liver disease. presence of ctDNA, and also characterise its
copy number aberration profile. Many existing
We analysed 3 publically-available WES
methods focus on using aberrant read-depth
datasets of HCC and paired non-tumour tissue
to infer the presence of copy number variation.
and generated our own datasets from donor
We have investigated the use of aberrant
liver tissue as a negative control. After filtering
allelic imbalance, as measured by the allele-
out probable germline mutations, we used two
specific read depths at SNP loci. We have
algorithms (PolyPhen and SIFT) to predict the
adapted methods which combine population
effect of the missense mutations on protein
level phasing with read-backed phasing in
function. We also determined whether these
order to maximise the information content in
SNPs were likely to alter genes expressed in the
allelic imbalance. We demonstrate that allelic
hepatocytes by filtering out those genes not
imbalance, modelled in this way, provides
detected in liver tissue based on our previous
sufficient power to detect tumour DNA even
data from microarray analysis.
when the mixture sample of tumour and normal
When more DPMs (but not benign SNPs) DNA at a minimum 5% tumour cellularity.
normalised to the total number of non- Moreover we are also able to pin down to
synonymous SNPs, significantly were found in the regions with copy number changes and
tumour tissue vs. non-tumour. This relationship predict possible aneuploidy.
was maintained when only liver-expressed
genes were analysed, suggesting that
damaging mutations affect cell phenotypes.
Accumulation was observed despite the
aetiology and known driver mutations in the
91
Handbook
2014 AGTA
Conference
The Mitogen Activated Protein Kinase (MAPK)

Poster 49 signaling cascade, which signals through
MEK and ERK proteins, has previously been
implicated in the regulation of EMT. In this
Erik Thompson, Melissa J Davis, Joseph Cursons,
work, we have used previously established
Karl Leuchowius, Mark Waltham, Sepideh
breast cancer derived cell line models of
Foroutan, Andrew Redfern, Eva Tomascovik-
epithelial-mesenchymal plasticity and applied
Crook, Bryce van Denderen, Tony Blick, Gayle
RNA sequencing to measure transcriptional
Phillip, Edmund Crampin and Ian Street
changes that occur as cells transition to a more
The role of MEK signaling in driving mesenchymal phenotype.
mesenchymal transition in cell line models We have identified co-ordinated changes in
of human breast cancer metastasis
the expression of genes in many, although
Abstract not all, MEK related signaling pathways, and
Epithelial mesenchymal transition (EMT) is the identify modulation in the abundance of key
process whereby sessile, polarised epithelial miRs implicated in regulating genes across
cells alter the expression of key adhesion and MEK related signaling pathways. Collectively,
regulatory molecules, and gain the ability our results demonstrate that different MEK
to survive and migrate as single cells. While related signaling pathways are altered by
this process occurs normally in development the induction of EMT in our cell line models,
we now recognize that metastasis has many depending on the position of the initial cell line
elements in common with developmental on the epithelial-mesenchymal spectrum, and
EMT, which is subverted by carcinoma cells the type of stimulus used to trigger transition.
to allow metastatic spread. Cancer cells This work was supported in part by the EMPathy
rarely undergo a full conversion to the National Collaborative Research Program
mesenchymal phenotype, and instead appear funded by the National Breast Cancer
to occupy a range of positions along a Foundation, Australia.
phenotypic spectrum between epithelial and
mesenchymal states.
The EMPathy Breast Cancer Network (BCN) is a
national collaborative effort including scientists,
surgeons, medical oncologists and a consumer
advocate investigating the role of EMP in
breast cancer recurrence. The 7 thematic
research projects of EMPathy BCN, including
the 9 program-funded ‘Satellite’ projects,
are aligned with the Cooperative Research
Centre for Cancer Therapeutics (CTx) (www.
cancercrc.com/index), so that any potential
drug targets identified may progress into the
CTx drug development program.
Multiple parallel approaches in the Target
Discovery theme were used to identify
candidate regulators and effectors of EMP.
A total of 10 functional or gene expression
experiments provided 7,950 significant events in
any one system, which were cross referenced
against 10 public breast cancer datasets
relevant to EMP and/or breast cancer stem
cells.
92
Handbook
2014 AGTA
Conference
Poster 50 Poster 51
Dr Vincent Corbin, Dr Clare Fedele, Ms Dr James Doecke, Dr Thierry Chekouo , Dr

Samantha Boyle, Dr Mark Shackleton, Assoc. Kimberly Glass, Dr Marieke Kuijjer, Professor Kim-
Prof. Anthony Papenfuss Anh Do
Deciphering genomic evolution in miRNA-target gene relationship analyses

heterogeneous tumours identify key oncogenes for Kidney Renal
Clear Cell Carcinoma progression
Abstract
Cancer genomes can be highly unstable. Abstract
Due to this instability, genomic evolution is Introduction: Although gene-cancer

typically rapid and highly heterogeneous relationships are easily identifiable and
within even a single tumour. The availability of usually strong, they are quite complex with
genomic analysis methods now allows for the multitudes of genomic elements responsible
investigation of cancer genome evolution. for the delineation of normal DNA replication,
Chromosomal SNP array and genome ultimately leading to cancer cell proliferation,
sequencing data provide information on and without treatment decreased survival time.
dominant and non-dominant alleles, and can Methodology: Using The Cancer Genome
be used to study heterogeneity by analysis of Atlas (TCGA) data repository, we used level
B allele frequency, log R ratio and nucleotide 3 Kidney Renal Clear Cell Carcinoma (KIRC)
mutation frequency. mRNA and miRNA expression data with the
We initially performed SNP array analysis on Passing Attributes between Networks for Data
a primary malignant melanoma sample. We Assimilation (PANDA version 2) method1. We
then derived single cells from the whole tumour compared biological networks identified
and propagated these in mouse xenografts. from two separate approaches: 1) network
Chromosomal SNP array analysis was then regulation via transcription factor (TF) – target
performed again on the tumour samples arising gene associations, and 2) network regulation
from the single cells. via miRNA - target gene associations.
Furthermore, we use PANDA to combine
Following comparison of the resulting copy
both miRNA and TF’s to assess the biological
number profiles, we were able to generate
networks where target gene regulation is
tumour-specific genograms, arising from the
present from both sources. We then assessed
primary tumour through its daughter tumours
these networks using the Ingenuity Pathway
and their subsequent heterogeneous genomic
Analyses (IPA) tool.
evolution. Comparison of copy number profiles
was performed using a novel method based Research design and hypotheses: Data was
on the T-statistic between sliding windows on grouped from tissue samples where the clinical
the log R ratio and B allele frequency channels. information indicated either FIGO stage 1/2,
or FIGO stage 3/4 for comparison, such that
Interrogation of the data arising from the
both TF-target gene and miRNA-target gene
tumour at multiple stages of evolution through
networks were available for comparison
the various daughter tumours illuminated
between cancer stages. Our research
specific evolutionary patterns, including the
hypothesis followed the premise that we would
temporal pattern of genomic rearrangements,
uncover differential networks for both TF-
key genomic variations which occur
target gene and miRNA-target gene networks,
independently within multiple daughter cell
between cancer stages. Moreover, we
lines, and the pattern of change in the rate of
aimed to discover novel complex interactions
tumour genome instability throughout multiple
between TF-miRNA and target gene related to
generations.
cancer progression.
93
Handbook
2014 AGTA
Conference
Data selection and pre-processing: From

a total of 20,532 mRNA’s, we selected only Poster 52
those genes with variance greater than the
median (plus one standard deviation), that
David Goode , Mandy Ballinger, Soroor Hediyeh
had less than 40% missing data, and that were
Zadeh, Paul Leo , Matthew Brown, Ajay Puri,
significantly associated with survival time,
Isabelle Ray-Coquard, The International
leaving a total of 1,387 mRNA markers. mRNA
Sarcoma Kindred Consortium, David Thomas
data was log transformed, QQ-normalized
and scaled prior to imputation using K-nearest Amplicon-based targeted resequencing
neighbours algorithm. From the 805 available identifies novel genetic risk factors for
miRNA markers, we selected those miRNAs with sarcoma and an association between
genetic load and age of onset in sarcoma
predicted target genes (target gene selection
patients
using the approach in 2) that matched the
1,387 mRNAs previously selected, leaving 458 Abstract
miRNA markers to add into the PANDA method. The risk of developing a sarcoma (a tumour of
Transcription factors (N=130) were selected bone, muscle or connective tissue) has a strong
using the list published in Glass et al., (2013). heritable component, yet our knowledge of
Results: Taking two separate approaches to sarcoma risk genes remains incomplete. This
gene-regulator selection, we initially selected study used the Haloplex amplicon sequencing
the top 50 genes (with their regulators) from assay to identify coding germline genetic
each of the comparisons (six in total) for variation in 100 targeted genes in 668 sarcoma
further network analyses with IPA. We then patients recruited through the International
subtracted the PANDA edge coefficient from Sarcoma Kindred Study (ISKS) who comprised a
the FIGO stage 1/2 group from the FIGO range of sarcoma subtypes and patient ages.
stage 3/4 group, taking the top 50 maximally SNVs and short indels were filtered on the
separated genes (and their regulators). basis variant frequency, predicted functional
Edge coefficients were greatest using the consequence, sequencing quality metrics and
PANDA method with miRNA markers alone annotations from the HGMDPro database.
(without TF’s), with a large proportion of strong Our filters achieved >95% specificity for SNVs
relationships identified. Comparison of edge and enriched for deleterious germline coding
coefficients between FIGO stage groups using variants, as evidenced by the association of
the miRNA data instead of TF’s also identified the presence of selected variants with clinical
the largest proportion of differential edges. relevant features, including early age of onset.
Of the gene-regulator relationships with the
Comparison to an internal set of 227 controls
strongest differences between cancer stage
sequenced using the same Haloplex capture
groups, we find 15 genes (CHRM1, CXCL5,
reagents identified genes known to be
CYP8B1, DSCAM, FAIM2, FUT3, GFAP, HAS2,
associated with sarcoma risk as well as several
IRF4, KCNMA1, MUC4, PEG10, PRDM16,
novel candidate sarcoma risk genes. These
PTPRZ1, SPIB) within a 30-gene network (Cell
findings were replicated two external exome
death and survival, cell cycle and cancer)
sequencing data sets from populations of
with IPA. Genes at the centre of this network
European heritage, using read depth to correct
included TP53, CDKN1A, NFkB, CXCL8, MYC
for coverage bias and synonymous variants to
and ERK1/2. Conclusions: Extending the PANDA
correct for ascertainment bias.
methodology to include miRNA-target gene
regulation elucidated multiple genes in cancer Furthermore, we uncovered evidence of
and cell death/survival networks. Future work positive epistatic interactions in germline
using this approach is aimed at identifying genetic variation, as patients carrying multiple
a pan-cancer pathway that may provide germline risk alleles had an increased risk of
therapeutic targets for many different cancers. developing sarcoma earlier in life.
94
Handbook
2014 AGTA
Conference
Our findings point to improvements in genetic and mRNA biomarkers. However, to date
screening for individuals at risk of early-onset there have been no such analyses applied
cancers and demonstrate the utility of germline to microRNA-based signatures of prognosis.
genetic variation in clinical decision-making. Our findings provide direct insights into the
translational potential of some signatures, while
highlighting critical limitations of others.
Poster 53
Poster 54
Ms Kaushala Jayawardana , Dr Sarah-Jane
Schramm , Dr Varsha Tembe, Dr Samuel Müller,
Professor Richard A. Scolyer, Professor Graham Anabel Kearney, Carrie Van Der Weyden &
J. Mann, Associate Professor Yee Hwa Yang Silvia Ling
MicroRNA prognostic signatures in Next generation sequencing of

metastatic melanoma haematological diseases in a clinical
setting a pilot study
Abstract
Predicting the survival outcome of patients is Abstract
an important context in many biological studies Molecular analysis of haematological

as it aids to identify sub groups of patients that diseases is of great benefit to clinicians and
would ultimately lead to the development patients alike not only in the stages of initial
of individualized treatment therapies. With diagnosis but also in the classification of risk
the recent advancements in high-throughput groups and prognostic outcome. This pilot
technologies and the generation of vast study examines the use of next generation
amounts of genetic information, the interest amplicon sequencing as a method of mutation
has been increasingly focused on using detection of five genes which have prognostic
these data or a combination of these data significance in acute myeloid leukaemia (AML)
to unveil intrinsic characteristics of patients and three genes involved in the diagnosis of
to aid in disease prognosis. In this context, myeloproliferative disease (MPD).
we analysed microRNA expression profiles of The three-fold aim of this study was to: i)
AJCC stage III melanoma patients. Outcome validate patient DNA panels, comparing the
in surgically resected stage III disease is mutation status resulting from our institute
highly uncertain with a limited number molecular lab testing; ii) detecting new
of reliable and validated biomarkers, mutations in patients not screened before
where the patient outcome is currently for particular MPD genes and in the form of
predicted using a limited set of clinical base substitutions in the AML panel where
and histological features. We identified the majority of mutations are detected and
a 12-microRNA signature for the association analysed by capillary electrophoresis, a
of longer survival versus shorter survival. method incapable of detecting such mutations
We assessed the prognostic utility of this and iii) challenge the present molecular testing
signature in relation to top-ranking AJCC scenario by detecting specific mutations for
clinico-pathologic prognostic indicators multiple diseases simultaneously using the same
and evaluated whether the signature adds sequencing cartridge.
predictive value when integrated with
Methodology followed the Illumina 16S
the same. Furthermore, we extended our
Metagenomic protocol for a 2 step PCR
study by conducting a formal and systematic
amplification method for amplicon and library
cross-validation of the microRNA-
preparation, substituting locus specific primers
based prognostic signatures in melanoma
targeting the genomic areas of interest for
to date. Previous studies have systematically
those pertaining to the 16S ribosomal RNA
reviewed and/or cross-validated protein
gene. The targeted regions in this study
95
Handbook
2014 AGTA
Conference
involved the FLT3, NPM1, CEBPA, WT1 and

Discriminant Analysis (sPLS-DA) identified a
IDH1 genes in cytogenetically normal (CN)
potential gene list for each LAF pattern and
individuals with AML and the JAK2, CALR and
correlated histology. Adjusted p-values were
MPL genes for patients suspected of MPD
obtained from limma.
without probable reactive cause. Paired end
Miseq target sequencing was performed sPLS-DA revealed multiple up-regulated
and subsequent results analysed using Miseq inflammatory and angiogenic related genes in
Reporter. LAF-BLANCHING (LB) lesions compared to NO-
LAF (NL) and LAF-NO-BLANCHING (LNB) cases
Targeted next generation sequencing is
including ICOS in OEH and TIE1 and PECAM1 in
highly applicable for mutation analysis in
OED lesions. Principal Component Analysis (PCA)
a clinical setting compared to standard
clustered OLP samples closer to OSCC likely due
laboratory procedures. It has the ability for
to inflammatory markers. Overexpression of
high throughput and the convenience and
CXCL13, FAM46C and ADAM28 were noted in
capacity of screening individuals with different
LNB samples across all lesions (OEH˂ OED˂ OLP˂
haematological diseases for pertinent,
OSCC). This study reveals a relative molecular
significant gene mutations in a single batch
profile of characterised oral lesions that not
workflow.
only helps to identify the molecular pathways
which may contribute to different phenotypes,
Poster 55 but may also assist in stratifying lesions at risk of
malignant transformation.
Farzaneh Kordbacheh, Nirav Bhatia, Camile S

Farah Poster 56
Molecular profile of oral mucosal

lesions characterised by direct tissue Mr Marcus Lefebure, Dr Richard Tothill, Dr Jason
autofluorescence Li, Dr Geoffrey Matthews, Dr Jake Shortt, Dr
Abstract Edwin Hawkins, Dr Elizabeth Kruse, Dr Micah
Loss of autofluorescence (LAF) using an Gearhart, Professor Vivan Bardwell, Professor
optical imaging system (VELscope™) has been Ricky Johnstone
associated with potentially malignant disorders Exome sequencing reveals Bcl6 co-
of the head and neck including oral epithelial repressor (Bcor) as a frequently mutated
dysplasia (OED) and squamous cell carcinoma tumor suppressor gene in E-Myc lymphoma
(OSCC). This study utilised whole transcriptome
Abstract
analysis of oral mucosal biopsies to determine
The Eμ-Myc mouse model employs a transgene
differential gene expression levels and their
mimicking the t(8;14) translocation found in
relation to LAF characteristics.
Burkitt’s lymphoma. Eμ-Myc mice develop
Whole transcriptome libraries were constructed Burkitt-like B-cell leukaemia/lymphoma with
from 43 frozen tissue biopsies with correlating 100% penetrance. Progression from pre-
LAF characteristics including 4 cases of OSCC, malignant to clonal lymphoid neoplasia in
10 cases of oral lichen planus (OLP), 15 cases the Eμ-Myc model ostensibly occurs when
of OED and 14 cases of benign oral epithelial secondary somatic mutations disrupt the p19ARF/
hyperplasia (OEH). Barcoded libraries were p53 tumour suppressor axis.
then sequenced using an Ion Proton (Life
We applied exome-sequencing to a cohort
TechnologiesTM, CA, USA). All reads were
of 18 spontaneous murine Eμ-Myc lymphomas
mapped to the human reference genome,
to determine novel somatic driver mutations
hg19, using Burrows-Wheeler Transform
capable of cooperating with Myc. In addition
(BWA) algorithm. Sparse Partial Least Squares
to exome-sequencing, we harvested blood
96
Handbook
2014 AGTA
Conference
samples throughout the tumour progression at

2-week intervals to allow a temporal acquisition Poster 57
analysis of mutations that were detected at
end-stage. Interestingly, concurrent driving
Dr Evangelina Lopez De Maturana, Dr Antoni
mutations were evident in clonal tumours
Picornell , Dr Manolis Kogevinas, Dr Mirari
from the sequenced cohort, with Cdkn2a
Marquez, Dr Alexandra Masson-Lecomte,
(encoding p19ARF) deletions co-occurring with
Dr Montserrat Garcia-Closas, Dr Nathaniel
an activating Kras mutation. In addition to
Rothman Dr Francisco X Real , Dr Michael
these mutations we identified novel protein-
Goddard , Dr Nuria Malats
truncating mutations in Bcl6 co-repressor
(Bcor), which are yet to be described in Eμ-Myc GENOME-WIDE MULTIMARKER MODEL TO PREDICT
lymphoma. RNAi mediated knockdown of Bcor THE RISK OF BLADDER CANCER RECURRENCE
in Eμ-Myc foetal liver haematopoetic stem cells
Abstract
reconstituted into congenic recipient mice
The aim of this study was to investigate if
accelerated lymphomagenesis, validating Bcor
genome-wide common SNP profiles identified
as a tumour suppressor gene in this model.
using Bayesian statistical learning improves
This study challenges the current paradigm the prediction of risk of recurrence in patients
of tumour progression in Eμ-Myc as the data with non-muscle invasive bladder cancer
indicate that this system is perhaps a ‘3-hit when added to the usual clinico-pathological
model’. BCOR loss-of-function mutations have prognosticators (CPP).
recently been identified in cytogenetically
We used data of 810 patients of the Spanish
normal acute myeloid leukaemia (CN-
Bladder Cancer Study, who were followed
AML), chronic lymphocytic leukaemia (CLL),
up during 10 years and genotyped using
retinoblastoma and medulloblastoma but
the Illumina Infinium HumanHap1M array.
it has not yet been validated as a genuine
The endpoint of interest was the time to first
tumour suppressor gene. This study shows that
recurrence (TFR). Predictive ability of three
Bcor loss-of-function mutations accelerate
models was evaluated: 1) including CPP, 2)
lymphomagenesis in a model of Burkitt-like
including only SNPs and 3) combining CPP and
leukaemia/B-ALL.
SNPs.
We used Bayesian sequential threshold (BST)
models to consider the time-to-event nature
of the data, as well as censoring data. To
accommodate and handle the large p small
n problem, we combined BST with LASSO.
Receiver operating characteristic curves with
area under the curve (AUC) and determination
coefficient estimates (R2probit) in a cross-
validation scenario were used to evaluate the
model predictions. Model with CPP classified
better the patients at risk of having a TFR
than the model with SNPs (0.62 vs. 0.55). In
addition, it explained a larger proportion of the
phenotypic variance than the model with SNPs
(3% vs 1%). Although adding SNPs to the CPP
improved the R2probit by 33%, it did not improve
the classification (AUC) of new patients.
Our study shows the difficulty in predicting TFR
using CPP and the little contribution of SNPs to
predict TFR.
97
Handbook
2014 AGTA
Conference
and prioritised using in silico methods, and

Poster 58 ultimately, in vitro techniques.
Results: DNA from over 100 samples has
Dr Mark McCabe, Dr Mark Cowley, Dr Sunita been captured for sequencing analysis. The
De Sousa ,Dr Melissa Vincent, Dr Reginald Lord, sequencing coverage was consistently high,
Dr Maija Corish, Dr Amber Johns, Professor with >90% of targeted regions sequenced to
Sandra O’Toole, Dr Ann McCormack, Associate >100x. Our panel detected 700 variants, and
Professor Marcel Dinger genomic data is currently being analysed.
Conclusions: Our cancer panel is novel in its
Stratification of the cancer genomic
landscape through the development of a applicability to various cancer types and is
pan-cancer genetic screening panel designed to uncover new, causative genes
as well as potentially identify alternative
Abstract treatment options.
Background and aims: Clinical use of cancer-
screening panels is increasing for limited
cancer types such as breast, lung and Poster 59
melanoma, due to their potential to identify
causative/contributory genetic aberrations
and the identification of new, or the Kelly Quek, Katia Nones, Ann-Marie Patch,
repurposing of existing therapeutics, leading Lynn Fink, Felicity Newell, David Miller, Karin
to better informed diagnoses with improved Kassahn, John Pearson, Nic Waddell, Sean
prognoses. The genetic basis underlying many Grimmond
rare, and/or poorly understood cancers,
Structural Variation in Pancreatic Cancer
remains ill-defined. Patients afflicted with these
Genomes
conditions still face limited treatment options
with poor prognoses. We have developed a Abstract
pan-cancer genetic screening panel, which One of the hallmarks of cancer is the
we aim to demonstrate the clinical utility for accumulation of genetic alterations which
stratifying the genomic landscape of disparate leads to uncontrolled growth of mutated cells.
cancer types, thereby improving treatment This alteration ranges from small mutations
options and prognoses through discovering to large structural rearrangements. Structural
new or repurposing existing therapeutics. variation includes duplications, deletions,
Panel development: Our custom-designed insertions, translocation and other more
gene panel through Roche/Nimblegen, complex genomics rearrangements. Despite a
combines 313 genes from various cancer large number of studies and rapid progress of
screening panels including Foundation genomic technologies, the characterization of
Medicine; enabling us to detect various structural rearrangements and their underlying
germline and somatic mutations. This resulted in formation mechanisms are yet to be fully
~0.9Mb of target sequence and should provide understood.
excellent coverage across our 4372 targeted Here we present a comprehensive
exons, after sequencing 24 samples in a single catalogue of somatic structural variation in
lane on an Illumina HiSeq 2500. 120 pancreatic primary tumours using high
Methods: DNA was extracted from breast coverage whole genome sequencing data.
(phyllodes), pancreas, pituitary, lung and Each cancer genome harbours a diverse
oesophagus tumour and/or whole blood array of somatic rearrangements and is
samples and subjected to library preparation different between individual genomes within
using the NEBNext Ultra kit. Sequencing data cancer type. This suggests that different
was aligned to the genome using bwa-mem, mechanisms may have present in the cancer
variants were identified using HaplotypeCaller, genome or repaired these rearrangements.
98
Handbook
2014 AGTA
Conference
Furthermore, we mapped the breakpoints to show very low methylation in non-neoplastic

base pair resolution and have characterized colorectal tissue and are candidate biomarkers
the sequence context of the breakpoint for stool-based assays, while 11 genes have very
junctions into five categories: blunt ends, low methylation in peripheral blood DNA and
microhomology, non-templated sequences, are candidate blood-based diagnostic markers.
genomics shards and free DNA associated In particular, two methylated gene biomarkers,
with mobile elements. Lastly, we described BCAT1 and IKZF1, looked most promising.
the patterns of joining broken chromosomal In a follow up study, this two-gene panel could
fragments in different subtypes of pancreatic detect 85 of 130 CRC cases in plasma specimens
cancer genomes to explore key biological from 2139 volunteers scheduled for colonoscopy,
insights to better understand the relationship which represents a sensitivity of 65%. In all non-
between driving forces and their targets. neoplastic cases (n=1283), a specificity of 94.9%
was observed.
Pilot release of the test commenced in July and
Poster 60
is the first ever commercially available blood test
for colorectal cancer in Australia.
Jason Ross, Ms Sue Mitchell , Dr Jason Ross , Dr
Horace Drew , Mr Thu Ho , Mr Glenn Brown , Dr
Graeme Young, Dr Trevor Lockett, Dr Susanne Poster 61
Pedersen, Dr Lawrence LaPointe, Dr Peter Molloy
Colorectal cancer methylation Dr Matthew Wakefield, Monique Topp, Valerie

biomarkers Heong, Alison Hadley, Australian Ovarian
Cancer Study, Elizabeth Swisher, David Bowtell,
Abstract
Clare Scott
The frequently methylated colorectal cancer
(CRC) biomarker genes SEPT9, VIM1 and TMEFF2, Selection Rather Than Mutation Drives
have been developed into diagnostic assays. Increased Platinum Resistance in a Patient
Despite their promise, there is considerable Derived Xenograft Model of Ovarian
potential for the development of new DNA Cancer
methylation biomarkers or panels to improve Abstract
the sensitivity and specificity of current CRC High Grade Serous ovarian cancer in clinical
detection tests. care is frequently treated with platinum based
To identify candidate genes methylated in a high therapy.
fraction of CRCs we examined gene reactivation Epithelial ovarian cancer response to platinum
in cell lines following 5-aza/TSA treatment and therapy is categorised as sensitive (a prolonged
also applied two novel in-house epigenome response), resistant (an initial response followed
methods, SuBLiME and Bis-Tag, for methylation by an increase in tumour size), or refractory
assessment in cell lines and clinical samples. We (increase in tumour size under therapy).
further evaluated differentially methylated genes
Ovarian cancer often becomes refractory under
using bisulfite amplicon deep-sequencing and
the selection of therapy, posing the question as
quantitative methylation specific PCR.
to whether the refractory phenotype is selection
Characterisaton of methylation in tumor, of a subset of tumour cells or due to new
adenoma and non-neoplastic colorectal acquired mutations.
tissue and healthy donor peripheral blood led
To investigate the role of selection versus new
to discovery of a panel of 23 genes showing
mutation we have assessed the clonal diversity
elevated DNA methylation in >50% of CRC tissue
in high grade serous ovarian cancer patient-
relative to non-neoplastic tissue. Six of these
derived xenografts. These xenografts form part
99
Handbook
2014 AGTA
Conference
of a consecutive cohort with characterised DNA

repair gene mutations and platinum response Poster 62
status that is reflective of the clinical outcome
of the patient (Topp et al 2014). Copy number
Dr Stephen Wong, Dr Andreas Behren, Dr
variation was determined by whole genome
Victoria Mar, Dr Katherine Woods, Dr Jason Li,
sequencing to assess clonal diversity and its
Dr Claire Martin, Dr Karen Sheppard, Professor
stability under serial xenografting and multiple
Rory Wolfe, Associate Professor John Kelly,
rounds of therapy. We find that clonal diversity
Professor Jonathan Cebon
is maintained in our model system, and is stable
in the absence of therapy. Under selection Whole exome sequencing identifies
by platinum therapy some clones develop a a recurrent RQCD1 P131L mutation in
refractory phenotype that is associated with a cutaneous melanoma
change in clonal diversity and prominence of Abstract
a pre-existing chromosomal rearrangement, Melanoma is often caused by mutations
supporting clonal selection rather than due to exposure to ultraviolet radiation. This
contemporary mutation as the driver of the study reports a recurrent somatic C>T change
resistant to refractory phenotypic change. causing a P131L mutation in the RQCD1
The selected regions indicate several (Required for Cell Differentiation1 Homolog)
candidate genes which we plan to pursue with gene identified through whole exome
a CRISPR screen. sequencing of 20 metastatic melanomas.
Topp et al. Mol Oncol. 2014 8:656-68. Further screening in 715 additional primary
melanomas revealed a prevalence of ~4%. This
represents the first reported recurrent mutation
in a member of the CCR4-NOT complex in
cancer. Compared to tumors without the
mutation, the P131L mutant positive tumors
were associated with increased thickness (p=
0.02), head and neck (p=0.009) and upper limb
(p=0.03) location, lentigo maligna melanoma
subtype (p=0.02) and BRAF V600K (p=0.04) but
not V600E or NRAS codon 61 mutations. There
was no association with nodal disease (p=0.3).
Mutually exclusive mutations of other members
of the CCR4-NOT complex were found in ~20%
of all subcutaneous melanomas from the TCGA
dataset suggesting the complex may play an
important role in melanoma biology. Mutant
RQCD1 was predicted to bind strongly to
HLA-A0201 and HLA-Cw3 MHC1 complexes and
may therefore act as a neoantigen conferring
a possible survival advantage to patients or be
susceptible for immunotherapies in this patient
subset.
100
Handbook
2014 AGTA
Conference
complexity of the analytical process, and the

Poster 63 constantly evolving nature of the research-
grade analytical components of the pipeline.
To this end we have adopted rigorous best
Dr Mark Cowley, Mr Kevin Ying, Dr Lisa Ewans,
practice software engineering approaches,
Ms Kerith-Rae Dias, Dr Warren Kaplan , Dr
including Agile scrum methodologies that grant
Timothy Furlong , Professor John Shine , Professor
us the flexibility to continuously improve our
Michael Buckley , Dr Tony Roscioli, Associate
processes and analytics.
Professor Marcel Dinger
To date, we have sequenced exomes from
Implementing Clinical Genomics for >200 patients, from a range of conditions,
diagnostics and research largely reflecting the undiagnosed caseload
Abstract
at the Sydney Children’s Hospital, as well as a
Next Generation Sequencing (NGS) has range of kidney disorders, osteoporosis, and
revolutionised genetic diagnostic testing, dilated cardiomyopathy. From the Sydney
substantially improving diagnostic yields in Children’s Hospital cohort, we performed
patients with Mendelian disorders, from 7-12%, whole exome sequencing on 122 individuals
to on average 25-30%. NGS is also having a from 53 families (mostly intellectual disability
positive impact on cancer stratification, non- and epileptic encephalopathy), and achieved
invasive prenatal testing, complex disease a 38% diagnostic yield, with an additional 28%
studies, and the emerging field of personal of cases with potential novel variants. Notably,
genomics. The intersection of medicine and early economic cost-benefit analyses are
genomics presents a unique opportunity for the demonstrating that whole exome sequencing
diagnostic test to be used to drive research, in the clinic is both cheaper and faster than
both for improving diagnostic yields, and traditional testing algorithms. For disorders
understanding the impact of genetic variation where the degree of locus heterogeneity
upon human health. Due to the substantial is small (ie <10 candidate genes), we have
cost of developing and clinically validating a experienced even higher diagnostic yield,
genomic test, we have taken the approach of such as in cases of thin basement membrane
focusing on whole exome and whole genome disease, X-linked Alport’s syndrome, and
sequencing, with targeted analysis over known Nail-patella syndrome. We have struggled
genes for diagnostic purposes, and then more to diagnose patients with polycystic kidney
exhaustive exome-wide analyses for gene and disease using whole exome sequencing, due
mutation discovery in undiagnosed cases. to pseudogenes confounding the alignment of
short reads to the major candidate gene, PKD1
In late 2012, we established the Kinghorn
(responsible for 85% of cases), which we are
Centre for Clinical Genomics (KCCG) with
attempting to resolve with long range PCR.
the aim of implementing clinical genomics
in Sydney. At the heart of the KCCG are I will present an overview of the systems
two Illumina HiSeq 2500 sequencers that are that we are developing to enable clinical
used for rapid turnover exome sequencing, genomics at significant scale, as well as some
and more recently, one the world’s first early success stories and challenges from
Illumina HiSeq X Ten sequencing systems, sequencing these clinical exomes and reflect
with capability of sequencing more than 300 on the possibilities presented by low-cost whole
whole human genomes per week. Much of genome sequencing in the clinic.
our work over the past 18 months has been
focused on building the infrastructure to
support robust, reproducible, efficient clinical
genomics pipelines, and working towards
NATA accreditation. Clinically accrediting
this process is challenging, in part due to the
101
Handbook
2014 AGTA
Conference
Poster 64 Poster 65
Dr Ruth MacKinnon, Dr Meaghan Wall, Gisela Mir, Tim Semple, Jason Li, Richard Tothill,
Associate Professor Lynda Campbell Andrew Fellowes
Genome characterisation of the cell Library preparation, targeted-hybridisation

line U937 provides further evidence for capture and sequencing of limiting
centromere capture and demonstrates the amounts of FFPE samples
use of multiple genome technologies to
examine evolution in cell lines Abstract
Massively-parallel (next-generation)
Abstract sequencing is now being rapidly adopted
We have carried out a detailed study of the for the detection of pathogenic mutations
genome of the myeloid cell line U937 using SNP in cancer patient samples. One of the great
(single nucleotide polymorphism) array with challenges faced in this process is the often
various FISH (fluorescence in situ hybridisation) limiting amounts and poor quality DNA that
techniques. Complex rearrangements were can be obtained from formalin-fixed and
defined using SNP arrays to determine copy paraffin-embedded (FFPE) tissues. With the
number aberrations and breakpoints, and steady improvements to DNA sequencing and
different FISH techniques to determine the the chemistry used for library preparation,
organisation of the aberrant segments there are now a number of commercial
within derivative chromosomes. While SNP options for processing and sequencing of FFPE
array identifies copy number variation and samples. However, these methods still require
breakpoints at high resolution, FISH can be used thorough testing before they can be used
to characterise intact chromosomes and single routinely in a clinical setting under strict quality
cells, and localise highly repetitive DNA regions. assurance guidelines. Targeted-hybridisation
One benefit of SNP array is that the B allele capture is routinely used in research and has
frequencies allow the two chromosome been established as one of the most cost-
homologues to be distinguished. For example, effective methods to explore large genomic
we showed that different chromosome 16 regions for diverse germ-line and somatic
homologues were involved in an abnormal variations. We therefore aimed to explore the
chromosome consisting of parts of chromosomes sensitivity, accuracy and reliability of different
4 and 16 and a chromosome consisting of 16p library preparation methods combined with
and 20p material. Similarly, the 1q material that hybridisation capture for sequencing of FFPE
had been translocated to chromosome 3 was samples from pathology archives.
not from the same homologue as the 1q material In this technical work we selected a good
that had been lost from a del(1q). quality reference sample (NA12878, HapMap)
This cell line provides an additional example and several tumor FFPE clinical samples to
of centromere capture, a recently described initially compare four library preparation
phenomenon in complex karyotypes. A short methods (SureSelect XT, Ovation SP Ultralow,
fragment containing the 11 centromere was Kapa and ThruPLEX-FD) and two capture
joined to chromosome 16 and 20 material, methods (SureSelect XT and NimbleGen
enabling formation of a viable chromosome. SeqCap EZ) at different starting amounts
By considering the organisation of a complex of DNA (20 ng to 500 ng). We used a 3 Mb
abnormal genome using a variety of techniques, custom panel designed to capture 600 genes
we have been able to make some inferences frequently mutated in cancer, some known
about how it has evolved. This strategy can fusions and one virus. Captured libraries were
help sort out evolutionary processes in cancer sequenced with an Illumina HiSeq2500 at 100
genomes with multiple rearrangement events. paired-end reads aiming a mean coverage
>500x.
102
Handbook
2014 AGTA
Conference
Although successful libraries were generated

from 20 ng of DNA, using 50 ng or less DNA Poster 66
resulted in low complexity libraries and
inefficient captures that produced high
Ms Mahtab Mirmomeni, Dr Kelly Wyres, Mr
duplication levels and/or low on-target
Benjamin Goudey, Dr Jeremy Wazny, Dr
when aligned to the targeted regions.
Andrew Mackinlay, Dr Priscilla Rogers, Dr Tom
Underperformance at these conditions was
Conway
also true for the good quality reference
sample. When starting with 100 ng of DNA A clinical ready system for scalable
duplication rates were still higher than those genomics analysis
expected in optimal situations but coverage
Abstract
was good for most of the targeted regions, with
With the rapid drop of the cost of genome
results very similar for the two capture platforms
sequencing, its adoption as a routine
tested; at 300 ng we reached standard
procedure in clinical settings is fast
performance QC metrics. The highest variability
approaching. It is essential to have platforms
in library efficiency between FFPE samples was
that can track and process the massive
irrespective of the library preparation method,
amount of data that will be produced. These
emphasizing the importance of a good sample
systems require mechanisms for management
QC to estimate capture and sequencing
of sequencing data, sample information
success at an early stage.
(metadata), audit trail, analysis, visualization
Variant calling and copy-number variation and higher-level analytics. In addition, they
analyses are being run to determine the need to be easy to use for lab technicians.
sensitivity of the method. We aim to increase Given that genome sequencing might be
the number of samples and run validation used in multiple different domains, such as
tests to establish the combination of wet- cancer genomics, pathogen genomics, etc, it
lab and bioinformatics that best suits the use is desirable to create systems that are flexible
of targeted-capture and high-throughput and not bound to a specific type of analysis.
sequencing for the study of FFPE samples in a We introduce, SEGP, a Scalable Enterprise
clinical setting. Genomics Platform, created to enable the
routine use of genome sequencing in clinical
settings. SEGP stores metadata as XML
documents in a database. This enables higher-
level analytics to be run on the data captured
in the system e.g. for detection of anomalies
or trends. All sequencing files are stored in a
file system outside the database. Metadata is
searchable. The result of running analyses on
metadata and the associated sequences are
captured as metaresult (human-readable and
searchable) and result records (raw outputs).
SEGP can easily adapted to different genomics
applications. All data types for a particular use
case, in addition to the workflows applicable to
those data types, are configurable outside the
platform using XML schema. This allows flexibility
to adapt the platform to different settings
without the need to modify source code.
103
Handbook
2014 AGTA
Conference
Poster 67 Poster 68
Mr John Pearson, Dr Ann-Marie Patch , Dr Katia Mr Simon Sadedin, Dr Alicia Oshlack

None, Dr Kelly Loffler, Dr Henry Tang, Dr Andrew
Challenges to CNV Detection in Clinical
Barbour, Dr Nicola Waddell , Professor Sean
Settings using Targeted High Throughput
Grimmond Sequencing Data
Next-generation Sequencing - Beyond Abstract
Variants
High Throughput Sequencing (HTS) is rapidly
Abstract gaining clinical acceptance as a cost effective
Next-generation sequencing (NSG) has had solution for diagnosis of highly penetrant but
a profound impact on genomics research genetically heterogeneous diseases. Despite
and is making inroads in the clinic as part of this success, the analysis methods applied are
the therapeutic decision making process. usually limited to detection of point mutations
While there are numerous classes of genomic and small insertions and deletions. Detection
variants that can be called from NGS data - of larger structural or copy number variants
including nucleotide substitutions, insertions, (CNVs) such as deletions, is often omitted
deletions, copy number changes and structural despite such variants being equally deleterious
variations – there are an increasing number and a common cause of many disorders.
of other classes of information that can be While many tools have been published in the
extracted from NGS data. These other classes literature for detecting structural variation
of information can also play a major role in in targeted sequencing, in practice, few
disease development and progression. The laboratories are implementing these in their
authors will discuss techniques and software clinical sequencing analysis pipelines.
tools (including a number developed by the In this work, we apply a novel simulation
authors) for looking at some of these including method to explore the performance of
telomere quantification, pathogen screening, CNV detection methods on real data. The
mitochondrial heteroplasmy and mutational simulation method inserts reads from real X
signatures. We will highlight the importance of chromosomes of male samples into female
these approaches by applying them to cancer samples to simulate single copy deletions. We
genomes from oesophageal and other cancer find that real world performance of published
types. CNV detection methods appears to be highly
sensitive to a range of factors including:
method-specific tuning parameters, the exact
sequencing technology in use, variability in
the quality of data, and the number and
type of samples compared in a batch. We
conclude that in order for HTS CNV detection
to become clinically accepted, methods must
be developed that can work in a highly robust,
self-calibrating fashion and that are well tuned
to popular sequencing and capture platforms.
Along these lines, we show results from our own
method, Angel, which is specialized to achieve
high accuracy and robustness on the HaloPlex
targeted sequencing platform.
104
Handbook
2014 AGTA
Conference
Poster 69 Poster 70
Dr Jo-Ann Stanton, Dr Chris Rawle, Ms Christy Dr Siamak Tafavogh, Associate Professor Daniel
Rand, Mr Chris Mason, Dr Richard Hall, Dr Catchpoole, Associate Professor Paul Kennedy
Jeremie Langlet, Dr Joanne Hewitt, Dr Hugo
Generating Actionable Knowledge from
Strydom, Dr Angela Todd
Complex Genomic Data for Personalised
A Handheld qPCR Device for Point-of-Care Clinical Decisions
and In-Field Diagnostics
Abstract
Abstract The human genome is complex. Within

We have invented a Handheld Quantitative this complexity lies information about the
PCR (HHD qPCR) device that can operate individual. In conventional pathology
on battery for up to six hours. The HHD clinicians group patients with similar cancer
qPCR device was tested using World Health aggressiveness grades into broad clusters and
Organization and IANZ accredited assays for deploy the same treatment regimens for all
E.coli STEC type strain NZRM3634, Influenza, patients in a cluster. For example therapy of
Adenovirus, Enterovirus, Norovirus genogroup patients with low aggressiveness grade may be
I & II, and Astrovirus. In side-by-side tests with minor surgery or constant observation.
larger laboratory based instruments and clinical The main issue with conventional pathology is
samples the HHD qPCR was comparable to, that deploying the same therapy for patients with
and in one case better than, in-laboratory the same aggressiveness grade might not lead
technology. Tests included measures for to the same outcome because each individual
sensitivity, precision, and inter-assay variability. has unique genotypic characteristics. Improved
The HHD qPCR detects both SYBR green and patient outcome requires the development of
FAM reporter dyes. We have used DNA and approaches that lead to personalised treatment
RNA as reaction templates. Probe-based and strategies.
intercalating dye assays have also been shown Computer Automated Diagnosis (CAD) links
to perform successfully on the HHD qPCR device. data mining and bioinformatics techniques to
The user interacts with the device via a tethered build a framework for personalized medicine.
connection to a laptop computer. In addition, We present a CAD that indicates the genomic
the HHD qPCR is wireless enabled to permit distances between patients, a concept we call
interaction with the operator via smart phone or ‘similarity-space’ which reflects the similarity or
tablet devices. Current reaction vessel volumes dissimilarity in patient genomic characteristics.
are equivalent to those found in a 96 well plate. It predicts the survival chance of new patients
Diagnosis of infectious disease in the field or at under the chosen therapy by computing the
the initial point-of-care brings the advantage of distance between the new and previously
rapid infection containment, accurate diagnosis treated patients.
and immediate selection of the appropriate The CAD must work with imbalanced and high
treatment. We see significant cost savings and dimensional datasets. Thus, a multi-step algorithm
benefits for the animal health sector with the is proposed and validated with preliminary
technology quickly migrating to use in human results. Step1: we use resampling technique to
health. For example, the HHD qPCR device balance the number of samples within different
could monitor for disease outbreaks as part of a patients classes; Step2, we develop an approach
national surveillance program. With the correct using bootstrapping and random-forest to
assay panel, our technology can determine the reduce the dimensionality; Step3, we develop a
disease-causing species or reveal antibiotic matrix-decomposition technique to map patients
resistance, all in real time either “cow-side” or in the similarity-space. Our paradigm is childhood
in the remote clinic. cancer and we aimed at testing the CAD in the
context of acute lymphoblastic leukaemia.
105
Handbook
2014 AGTA
Conference
Poster 71 Poster 72
Ms Kim Ton, Macmil, S, Miller, A.L,Chua, E.W, Dr Margaret Jordan, Dr Louise Laverick , Dr
Rand, C, Stanton, J.A, Kennedy, M.A Melissa Gresle, Dr Dragana Stanley, Ms Letitia
Smith, Dr Judith Field, Dr Laura Johnson, Dr Tim
Next generation sequencing for
Spelman, Professor Helmut Butzkueven, Professor
pharmacogenetic screening
Alan Baxter
Abstract
Causes of Multiple Sclerosis: a functional
Pharmacogenetics (PGx) is contributing to
genomics approach
the improvement of individualized medication
and the prediction of clinical treatment Abstract
outcome. With advances in this field, a Multiple Sclerosis (MS) is the most common
wide range of PGx assays have been being disabling neurological disease affecting young
developed and launched, ranging from adults in Western Society. Both environmental
low-throughput approaches that genotype and genetic associations have been confirmed.
single genetic mutations to high-throughput To date, 110 strongly associated single
approaches to evaluate hundreds of genes nucleotide polymorphisms (SNPs) have been
simultaneously. The rapid evolution of next discovered but these represent common
generation sequencing (NGS) makes it possible human genetic variations, and their individual
to sequence a large number of genes, up to association with MS is weak. The key challenge
the complete genome, in a single reaction. therefore is to identify the causal genes. One
Hi-plex is an approach for generating NGS clue lies in the fact that many SNPs are in
libraries via single multiplex PCR amplification close proximity to known immune genes that
(Nguyen-Dumont et al. 2014). This method could affect changes in immune cell function
uses automated primer design software and by changing their RNA expression and their
a simple, low-cost protocol for cost-effective encoded proteins. As it’s highly likely that eQTLs
and rapid performance. We are interested in and splice QTLs will be cell subset specific we
exploring the application of next generation wished to analyse expression levels in individual
DNA sequencing methods to understand the cell subtypes. While T cells likely play a critical
impact of pharmacogenetic variation on drug role in MS pathogenesis, successful drug trials
responses and risk of adverse drug reaction. In to date have targeted B-cells and monocytes
this study, we use this the Hi-plex approach to in addition to T cells and there is also strong
develop an throughput assay for genotyping clinical and in vitro data identifying NK cell
pharmacogenetic variants of CYP2D6, deficiencies in MS patients. We thus examined
CYP2C9, CYP2C19, DPYD, TMPT, UGT1A1, gene expression changes in 5 cell subtypes
SLCO1B1, VKORC1. These genes are involved (CD4 and CD8 T cells, B cells, monocyes and NK
in metabolism of a wide range of drugs, and cells) using Affymetrix Human Gene 1.0 ST arrays.
some are also implicated in adverse drug Expression values were standardized across
reactions. We have focused on those genes chips using RMA and quantile normalization and
with the best evidence of clinical utility (refer genes were ranked by expression difference
to PharmGKB VIP). Initially we are developing significance by MWU and ANOVA. Expression
this assay for the Ion Torrent PGM platform. differences of those genes adjacent to MS
Our current iteration of this assay contains 15 associated risk SNPs were considered. An
amplicons for these eight genes, with indexing investigation of candidate genes identified and
for 12 patients samples, and we aim to expand the pathways to which they contribute is being
this further for more routine application. undertaken to allow more targeted therapies to
Reference: Nguyen-Dumont, T., Pope, B. J., Hammet, be developed.
F., Southey, M. C., & Park, D. J. (2013). A high-plex PCR
approach for massively parallel sequencing. BioTechniques,
55(2), 69–74. doi:10.2144/000114052
106
Handbook
2014 AGTA
Conference
Analysis of this data has revealed the temporal

Poster 73 expression pattern of HSA21 genes, with some
genes escaping the dosage bias at some
time points and not others. Additionally, non-
Nicholas Matigian, Elizabeth Mason, Dmitry
HSA21 genes display dysregulation at different
Ovchinnikov, Othmar Korn, Ernst Wolvetang,
stages of development. We conclude that the
Christine Wells
effects of trisomy21 manifest in two forms (1)
How does chromosome 21 trisomy alter dosage of HSA21 elements that directly alter
the gene regulatory network of neural the phenotype of maturing neuronal cells, and
differentiation? (2) temporal modulation of HSA21 elements
that directly or indirectly alter the normal
Abstract
developmental trajectory.
Down syndrome is a complex developmental
and neurodegenerative disorder caused by
trisomy of human chromosome 21 (HSA21). It
Poster 74
occurs in approximately 1 in 700 live births, and
is characterised by a hallmark facial features,
growth delays, mental retardation and
Associate Professor Leanne Dibbens, Michael G
premature neural aging, heart defects as well
Ricos, Sarah E Heron, Bree L Hodgson, Ingrid E
as a range of other variable symptoms. These
Scheffer, Samuel F Berkovic
are assumed to arise from increased dosage
of the HSA21 gene products, but this has not Epilepsy: new genes and old pathways in a
previously been modelled in the context of complex disorder
a developmental hierarchy of transcriptional
Abstract
regulation.
Epilepsy is the most common serious
Using iPSc generated from two DS and neurological disorder affecting 2% of the world
two control patients we modelled early population. More than 30 different epilepsy
development of neurons. We stepwise syndromes have been defined based on
differentiated these iPSc (Day0) into neuro- various clinical criteria. Around 25% of cases
ectodermal cells (day6), neural stem cells are symptomatic and are associated with due
(day12) and early neuronal progenitors to brain insults eg. stroke, tumour or brain injury
(day 18) using an established neuronal but 70% of all epilepsies have a genetic cause
differentiation protocols. At each stage making them ideal for molecular genetic
we performed Capped Ananlysis of Gene analyses. Over the past 15 years our group has
Expression (CAGE) RNA-seq as part of the collected more than 10,000 epilepsy cases
FANTOM5 project. This method produces gene from around the world with linked clinical
expression profiles of all genes coupled with information into a bio-bank. We have been
transcription start sites (TSS) position. analyzing this cohort with evolving genomic
This work will identify key regulators in neuron technologies and have identified a number of
differentiation that are sensitive to HSA21 genes as the cause of “monogenic” epilepsies.
trisomy altering the gene regulatory network Where available we have used linkage analysis
of DS iPSc. The project uses a modular network and next generation sequencing approaches
approach to investigate the transcription to identify candidate genes in epilepsy and its
factor network of in vitro iPSC->neural co-morbidities including intellectual disability,
differentiation. Perturbation of the network autism spectrum disorders and psychiatric
imposed by Chr21 dosage will be assessed features.
in terms of network membership, degree, or A number of ion channels genes including
kinetic shifts, with the aim of predicting the nicotinic acetylcholine receptors (e.g.
molecular drivers of cellular defects that CHRNA4), and sodium gated potassium
contribute to Down Syndrome pathologies. channels (e.g. KCNT1) have formed the
107
Handbook
2014 AGTA
Conference
majority of “epilepsy genes” discovered. We correlated to extensive metabolic phenotypic

have identified genes involved in secretory measures of insulin sensitivity in humans.
pathways (e.g. PRRT2, GOSR2, SCARB2). Inhibiting these proteins impaired insulin-
Interestingly we have also Identified structural stimulated glucose uptake in L6 myotubes,
proteins (e.g. PCDH19). Despite the larger confirming their role in insulin action. These
number of ‘epilepsy’ genes identified these data indicate the potential utility of using
only account for a small proportion of the systems biology approaches to segregate
cases of epilepsy studied. Significantly, we individuals early based on differential diabetes
have now identified signal transduction risk.
molecules from the mTOR pathway in focal
epilepsy, finding DEPDC5 mutations in 12% of
cases of familial focal epilepsy studied, making Poster 76
it the most common known cause of epilepsy.
Nicole Cloonan, Hilary C Martin, Shivangi Wani,

Poster 75 Anita L Steptoe, Keerthana Krishnan, Katia
Nones, Ehsan Nourbakhsh, Alexander Vlassov,
Sean M Grimmond
Dr Rima Chaudhuri, Dr Poh Khoo, Dr Jean Yang,
Imperfect centered miRNA binding sites are
Dr Jagath Junutula, Dr Ganesh Kolumam, Dr
common and functional
Zora Modrusan, Dr Kyle Hoehn, Dr Samantha
Hocking, Dr Jerry Greenfield, Professor David Abstract
James MicroRNAs (miRNAs) bind to mRNAs and
target them for translational inhibition or
Cross-species Systems Analysis of a
transcriptional degradation. It is thought that
Metabolic Disease to Find A Gene Signature
Predictive of Insulin Resistance most miRNA-mRNA interactions involve the
seed region at the 5′ end of the miRNA.
Abstract The importance of seed sites is supported
Insulin resistance (IR) is perhaps the best by experimental evidence, although there
predictor of future development of type 2 is growing interest in interactions mediated
diabetes (T2D). Identification of molecular by the central region of the miRNA, termed
signatures that can identify individuals with centered sites. To investigate the prevalence of
higher risk of developing T2D could enable these interactions, we apply a biotin pull-down
early stage intervention. Gene expression data method to determine the direct targets of ten
could provide an ideal tool for identification human miRNAs, including four isomiRs that
of such a molecular signature, but human share centered sites, but not seeds, with their
gene expression (GE) data is inherently noisy canonical partner miRNAs.
and highly variable. To overcome the lack
We confirm that miRNAs and their isomiRs
of power, we integrated human data with
can interact with hundreds of mRNAs, and
more stable data from a model organism and
that imperfect centered sites are common
thereby developed a cross-species (human-
mediators of miRNA-mRNA interactions. We
mouse) analysis platform. We applied this new
experimentally demonstrate that these sites
approach to identify a gene signature that
can repress mRNA activity, typically through
could differentiate insulin resistant from insulin
translational repression, and are enriched
sensitive individuals with improved prediction
in regions of the transcriptome bound by
accuracy (~20% increase) compared to
AGO. Finally, we show that the identification
standard clinical measures (OCM). This
of imperfect centered sites is unlikely to be
signature also identified beta-catenin and
an artifact of our protocol caused by the
JAK1 as novel ‘connection hubs’ to the Insulin
biotinylation of the miRNA. However, the fact
Signaling Pathway whose GE pattern highly
that there was a slight bias against seed sites in
108
Handbook
2014 AGTA
Conference
our protocol may have inflated the apparent on a collection of 86 M. tb clinical isolates from
prevalence of centered site-mediated South Africa, sequenced as part of another
interactions. study. We show that the lineages that the
Our results suggest that centered site-mediated spoligotypes associate to are consistent with
interactions are much more frequent than a phylogenetic tree derived from variation
previously thought. This may explain the across the whole genome. Furthermore,
evolutionary conservation of the central region spoligotype-defined lineage assignments
of miRNAs, and has significant implications for were consistent with those known for publicly
decoding miRNA-regulated genetic networks, available genome sequences included in the
and for predicting the functional effect of phylogenetic analysis. Our method is efficient
variants that do not alter protein sequence. and robust, meaning it is appropriate for
research deployment and public health use.
Poster 77
Poster 78
Dr Thomas Conway, Dr Kelly Wyres, Dr Keira

Cohen, Dr Tal El-Hay, Dr Omer Weissbrod, Dr Harriet Dashnow, Susan Tan, Dr Debjani Das,
William Bishai, Dr Alex Pym Professor Simon Easteal, Dr Alicia Oshlack
Deriving Mycobacterium tuberculosis Genotyping microsatellites in next-

Spoligotypes from NGS Data generation sequencing data
Abstract
Abstract
Spoligotyping is a genotyping assay Microsatellites are short (2-6bp) DNA
applicable for Mycobacterium tuberculosis sequences repeated in tandem, which make
(M. tb), the causative agent of tuberculosis. up approximately 3% of the human genome.
It is used for identifying the relationships These loci are prone to frequent mutations and
between isolates, tracking the spread and high polymorphism. Dozens of neurological and
evolution of M. tb. Traditional spoligotyping developmental disorders have been attributed
is a laboratory technique, whereby genomic to microsatellite expansions. Microsatellites
DNA is assayed for hybridisation to a set of have also been implicated in a range of
43 nucleotide probes, each 25 base pairs in functions such as DNA replication and repair,
length. Researchers are increasingly moving chromatin organisation and regulation of gene
away from traditional genotyping techniques expression.
in favour of high resolution next-generation Traditionally, microsatellite variation has been
sequencing (NGS), which is also generating measured using capillary gel electrophoresis.
much interest in the clinical and public health In addition to being time-consuming, and
microbiology space. However, derivation of expensive, this method fails to reveal the full
traditional genotyping information from these complexity at these loci because it cannot
data remains key for backwards compatibility detect SNP polymorphisms and compound
and contextualisation of results. microsatellites.
We present a method for deriving spoligotypes Next-generation sequencing has the potential
from NGS data. We use an alignment-free to address these problems. However,
method that avoids the cost and biases determining microsatellite lengths using next-
associated with mapping or de novo assembly, generation sequencing data is difficult. In
and results in an assay that closely mirrors the particular, polymerase slippage during PCR
original hybridization based assay. Isolates for amplification introduces stutter noise. A small
which experimental spoligotypes and NGS number of software tools claim to genotype
data are both present were not available, so simple microsatellites in next-generation
we demonstrate the efficacy of the method sequencing data, however they fail to address
109
Handbook
2014 AGTA
Conference
the issues of SNPs and compound repeats, genes associated with a disease often
and they tend to provide only approximate interact. Thus, we propose a novel kernel that
genotypes. incorporates the topology of pathways and
We have developed a microsatellite information on interactions. Using simulation
genotyping algorithm that addresses these studies, we demonstrate that the proposed
issues, providing high accuracy as well as method maintains the type I error correctly
more detailed analysis of microsatellite and can be more effective in the identification
loci. We have validated it using high depth of pathways associated with a disease than
amplicon sequencing data of microsatellites non-network-based methods. We apply our
near the AVPR1A gene. We found high approach to genome-wide association case-
concordance between our algorithm and control data on lung cancer and rheumatoid
repeat lengths obtained by electrophoresis, arthritis. We identify some promising new
manual inspection and Mendelian inheritance. pathways associated with these diseases,
By subsampling the reads, we found that our which may improve our current understanding
model is accurate to within one repeat unit of the genetic mechanisms.
down to coverages that we would expect in
standard exome sequencing.
Poster 80
Poster 79
Miriam Fanjul Fernandez, Peter Hickey, Vesna
Lukic, Natasha Brown, Greta Gillies, Sarah
Dr Saskia Freytag, Dr Juliane Manitz, Professor Wilson, Martin Delatycki, Ingrid Scheffer,
Dr Martin Schlather, Professor Dr Thomas Melanie Bahlo, Paul Lockhart
Kneib, Professor Dr Christopher Amo, Professor Gene discovery in large multiplex families
Dr Angela Risch Dr Jenny Chang-Claude, with autism spectrum disorders
Professor Dr Joachim Heinrich, Professor Dr
Heike BickebÃ¶ller Abstract
Autism spectrum disorders (ASD) are common
Identifying disease-associated pathways neuro-developmental disorders, defined
in genome-wide association studies via a by impairment in language and social
network-based kernel approach
interaction, that affect ~1:100 individuals. ASD
Abstract typically display complex inheritance with a
Biological pathways provide rich information multifactorial basis, but despite the significant
and biological context on the genetic causes heritability, the genetic basis underpinning the
of complex diseases. The logistic kernel disease remains largely unexplained. Moreover,
machine test integrates prior knowledge up to 20% of relatives of individuals with ASD
on pathways in order to analyze data display the “broader autism phenotype” (BAP),
from genome-wide association studies characterised by less severe deficits in one or
(GWAS). In this study, the kernel converts the more of the three core ASD domains.
genomic information of two individuals into Genetic studies have demonstrated that next-
a quantitative value reflecting their genetic generation sequencing (NGS) technology
similarity. With the selection of the kernel, one can be a powerful tool to identify the
implicitly chooses a genetic effect model. Like causative genes. However, the high genetic
many other pathway methods, none of the heterogeneity characterising ASD along with
standard kernels accounts for the topological studies focusing on small multiplex kindreds
structure of the pathway or gene-gene may explain previous unsuccessful molecular
interaction types. However, there is evidence studies. We have employed a novel approach
that connectivity and neighborhood of genes that analyses the BAP as marker of an ASD
are crucial in the context of GWAS, because genetic variant in large multiplex families that
110
Handbook
2014 AGTA
Conference
are more likely to be genetically homogenous mutations in seventeen different disease

and suitable for classical linkage approaches genes and a mitochondrial DNA deletion were
to gene discovery. To date, more than 40 identified. These mutations were inherited in
individuals from two large pedigrees have maternal (43%), autosomal recessive (45%)
been exome sequenced and genotyped using or X-linked (12%) manners. The identified
Illumina platforms. Reads were aligned to the disease genes encoded subunits or assembly
reference genome (GRCh37) using Novoalign factors of oxidative phosphorylation (OXPHOS)
and single point mutations and indels were complexes I (35%), IV (23%), V (20%) or pyruvate
called simultaneously across all sample BAMs dehydrogenase (12%). Molecular defects in
with GATK-Haplotype Caller and annotated genes associated with a combined OXPHOS
with Annovar. On average, individuals have deficiency were also observed in 10% of patients.
300,000 raw variants, including artefacts and We recently characterised PET100 as a novel LS
non-relevant variants. After applying numerous disease gene encoding a complex IV biogenesis
filtering steps based on allele frequency, factor. A pathogenic homozygous c.3G>C
biological consequences and inheritance mutation predicted to abolish the initiation
models, we have selected several candidate codon was identified in ten patients (8 families)
genes for further evaluation. Finally, analysis of of Lebanese descent. Lentiviral-mediated
the segregation of the mutation within each overexpression in patient fibroblasts restored
family and functional studies will allow us to complex IV assembly and activity, confirming it
identify the causative ASD variant. as a bona fide disease gene. Our results highlight
the significant genetic heterogeneity underlying
LS. Future MPS based studies are anticipated to
Poster 81 reveal additional novel LS disease genes.
Nicole Lake, Sze Chern Lim, Katherine R. Smith, Poster 82

David A. Stroud, Alison G. Compton , Luke
C. Gandolfo , John Christodoulou, Michael T.
Ryan, Melanie Bahlo, David R. Thorburn Victoria M. Perreau, Alberto Capurro, Liviu
Bodea, Patrick Schaefer, Ruth Luthi-Carter
The genetic basis of Leigh syndrome, the
most common mitochondrial disorder Computational deconvolution of genome
affecting children: A cohort study wide expression data from Parkinson’s
and Huntington’s disease tissues using
Abstract Population-Specific Expression Analysis.
Leigh (-like) syndrome (LS) is the most common
paediatric clinical presentation of inherited Abstract
mitochondrial energy generation disorders. The characterization of molecular changes

This progressive neurodegenerative disease is in diseased tissues gives insight into
characterised by the development of bilateral pathophysiological mechanisms and is
symmetrical lesions in the brainstem and basal important for the development of treatments.
ganglia. However LS is clinically and genetically Genome-wide gene expression analysis has
heterogenous, and can be caused by mutations proven valuable for identifying biological
in over 50 different genes. We present a cohort processes in neurodegenerative diseases
of 113 patients (95 pedigrees) with a clinical, using post mortem human brain tissue and
radiological and genetic diagnosis of LS. numerous datasets are publically available.
Mutations were identified through diagnostic However, many studies utilize heterogeneous
screening or targeted massively parallel tissue samples consisting of multiple cell
sequencing (MPS) research studies. Presentation types, all of which contribute to global gene
age was available for >80% of patients, with a expression values, confounding biological
median age of onset of 8 months. Pathogenic interpretation of the data. Elucidation of the
111
Handbook
2014 AGTA
Conference
cell-type-specific contributions to expression detected by MRI analysis. The focal epilepsy

profiles enables better understanding of the showed autosomal dominant inheritance and
underlying processes. For example, changes in linkage analysis failed to identify any linkage
numbers of neuronal and glial cells occurring region.
in neurodegeneration confound transcriptomic We carried out exome sequencing on two
analyses, particularly in human brain tissues individuals from the family who were affected
where sample availability and control is limited. with focal epilepsy and displayed abnormal
To this aim, we have applied our recently MRI. The sequence data was analysed using
published computational deconvolution an in-house bioinformatic pipeline. Candidate
method, population specific expression causative genetic variants were identified and
analysis (PSEA). PSEA estimates cell-type- validated by Sanger sequencing. These were
specific expression values using reference analysed for co-segregation with affected
expression measures, which in the case status and assessed for likely pathogenicity.
of brain tissue comprises mRNAs with We identified a mutation in DEPDC5 as being
cell-type-specific expression in neurons, causative of both lesional and non-lesional
astrocytes, olygodendrocytes and microglia. focal epilepsy in the family. We also identified
As an exercise in PSEA implementation two other families with DEPDC5 mutations
and hypotheses development regarding who also had mutation-positive individuals
neurodegenerative diseases, we applied PSEA with lesional and non-lesional focal epilepsy.
to Parkinson’s and Huntington’s disease (PD, DEPDC5-associated malformations included
HD) datasets. Genes identified as differentially bottom-of-the-sulcus dysplasia (3 members
expressed in specific cell types by PSEA from 2 families) and focal subcortical band
were validated using external data. Network heterotopia (1 individual). We show that
analysis of differentially expressed genes and mutations in DEPDC5 cause familial cases
Annotation Clustering (DAVID) elucidated of focal epilepsy associated with structural
molecular processes implicated by differential lesions. Previously we found that mutations in
gene expression in specific cell types. The DEPDC5 caused familial cases of non-lesional
results of these analyses provided new insights focal epilepsy. We therefore now show that
into the implementation of PSEA in brain lesional and non-lesional epilepsy can have
tissues and additional refinement of molecular a shared genetic aetiology. This challenges
signatures in human HD and PD. previous dogma of lesional and non-lesional
epilepsy being regarded as distinct entities.
DEPDC5 is part of the GATOR1 complex, which
Poster 83 negatively regulates mTORC1 signalling and
DEPDC5 mutations are associated with brain
malformations similar to Tuberous Sclerosis
Dr Michael Ricos , Dr Sarah Heron, Dr Ingrid
patients where TSCI and TSCII are mutated.
Scheffer, Ms Bree Hodgson, Dr Simone
Mandelstam , Dr Douglas Crompton, Dr
Francesca Bisulli, Dr Paolo Tuniper, Professor
Samuel Berkovic, Associate Professor Leanne
Dibbens
MUTATIONS IN THE mtorc1 REGULATOR DEPDC5

ARE A MAJOR CAUSE OF LESIONAL AND NON-
LESIONAL FOCAL EPILEPSY
Abstract
We set out to find the genetic cause of focal
epilepsy in a family with individuals that had
lesional and non-lesional focal epilepsy, as
112
Handbook
2014 AGTA
Conference
Poster 84 Poster 85
Lyndal Henden, Melanie Bahlo, Marie Shaw, Ms Tracey Van Stijn, Ms Shannon Clarke, Mr
Alison Gardner, Jozef Gecz Rudiger Brauning, Mr Ken Dodds, Mr John
McEwan
Inferring Identity by Descent on the X
Chromosome Genotyping-by-Sequencing (GBS):
Comparison to SNP chips and Whole
Abstract Genome Sequencing
Identity by descent (IBD) detection has played
a crucial role within the field of human disease Abstract
mapping. The idea behind IBD in disease Genotyping-by-Sequencing (GBS) has the
mapping is that if a disease is heritable, then potential to be a cost effective, reproducible
affected individuals in the same family are and high-throughput SNP genotyping method.
likely to have the same disease-causing Although we have been investigating GBS in a
variant. Affected individuals will then also number of livestock and aquaculture species,
have a shared region of IBD around the causal we have concentrated largely on sheep. To
variant that will not be present in unaffected establish the accuracy and reproducibility
individuals. Substantial progress has been of GBS, and to allow forward genotyping
made with regards to IBD methodology such capability, SNPs located in GBS targeted
as the inclusion of genotyping errors and genome regions were included on the sheep
accounting for linkage-disequilibrium, driven by High Density (HD) Illumina SNP chip. We have
the generation of denser marker data such as compared genotype calls between the two
that generated by next generation sequencing genotyping methods as well as whole genome
(NGS), however little to no work has been done sequencing (between 10x – 20x genome
on the X-chromosome. Our work focuses on the coverage).
“forgotten X” and as such we have developed
a Hidden Markov Model for inferring IBD sharing
on the X-chromosome. This method is designed
for SNP array genotype data, however it can
also be applied to NGS data once the data has
been formatted as SNP array data. Using exome
sequencing data from individuals with X-linked
intellectual disability, we show that our method
can narrow the search space for causal variants
on the X-chromosome as well as consolidate
evidence that a variant is causal.
113
Handbook
2014 AGTA
Conference
Poster 86 Poster 87
Ms Anita Goldinger , Dr Jian Yang, Dr Anjali Ms Beth Signal, Dr Brian Gloss, Mr Dominik
Henders, Dr Allan McRae, Professor Grant Kaczorowski, Mr Seth Cheetham, Ms Franziska
Montgomery, Professor Peter Visscher, Professor Gruhl, Dr Marcel Dinger
Greg Gibson, Dr Joseph Powell
The discovery of novel lincRNA with
Extensive genetic co-regulation drives potential roles in mESC development
highly correlated modular structures in
transcriptomic data Abstract
A large proportion of the genome can be
Abstract transcribed as long non-coding RNA (lncRNA),
Investigating the genetic co-regulation of RNA a class of non-coding RNA with demonstrated
transcripts can elucidate the complex genetic biological importance. Despite multiple
architecture governing the translation of gene available computational tools being available,
products into higher-order phenotypes. In a the categorisation of these transcripts as
previous study, we identified nine modules of non-coding from sequence information alone
correlated genes, which were representative remains difficult.
of distinct changes in immunological function
due to infection or geographical location. Here, It is also highly likely that novel lncRNAs
we report results from our examination into the remain to be found due to their characteristic
genetic co-regulation of these modules. The expression pattern of low levels combined with
phenotypic correlations between module probes high cell-type specificity and rapid turnover.
(ranging from 0.35 to 0.89) are significantly We hypothesised that employing higher
higher than estimates obtained from randomly temporal resolution RNA-Seq time courses
sampled probes. The probe heritability (h2), would allow us to capture novel, transiently
estimated using GREML analysis, was high with expressed or highly specific, lncRNA transcripts.
> 90% of the probes demonstrating a significant
h2 between 0.19 and 0.65. To calculate the We thus sought to identify potentially novel
amount of covariance attributed to common intergenic lncRNAs (lincRNAs) important
genomic control, we calculated the genetic in cellular differentiation using RNA-Seq
correlation (rG) using bi-variate GREML, for all data from a six-hourly time course over six
pair-wise combinations of module probes. The days of mouse embryonic stem cell (mESC)
average rG estimate of these probes was higher development. Genome guided transcriptome
than the phenotypic correlations, ranging from reconstruction was employed to generate a
0.56 to 0.96. The high rG suggests that common ‘novel transcriptome’ of mESC differentiation.
loci regulating module probes are driving the Transcripts from this were then subjected to
high phenotypic correlations. A significant a lincRNA detection pipeline established
number of trans-eQTLs were shared between by comparing the performance of multiple
probes in each module. These eQTLs were computational tools for classifying transcripts
located in genes involved in hematopoiesis as coding or non-coding on annotated
and were also enriched for previous GWAS hits mouse transcripts. This method facilitated
with various blood cell related disease. Our the discovery of hundreds of novel lincRNA
results demonstrate that these modules are with potential roles in mESC development.
under tight genetic regulation, which facilitates Insights into potential functions of these
their coordinated response to environmental novel transcripts were gained by comparing
pressures. This information has important expression profiles to those of known transcripts
application in the functional annotation of and detecting expression changes in other
genomic loci, building causal networks driving developmental transcriptome datasets.
disease and exploring the basis of disease
susceptibility between individuals.
114
Handbook
2014 AGTA
Conference
scope to diversify innate immune responses

Poster 88 by producing variant protein products, and
impacting post-transcriptional transcript fate.
These findings indicate TSS engagement
Suzanne Butcher, Edward Huang, Ashley
patterns are an exciting regulatory layer which
Waardenberg, Othmar Korn, Rowland
should be considered when deciphering
Mosbergen, Tyrone Chen, Kelly Hitchens, Dipti
phenotypic diversity in innate immune
Vijayan, Anthony Beckhouse, Mark Walker, Tim
responses.
Barnett, James Frazer, Antje Blumenthal, Matt
Sweet, The FANTOM5 consortium and Christine
Wells
Poster 89
Diversification of innate immune responses
by transcriptional mechanisms
Dr Brian Gloss, Ms Bethany Signal, Mr Dominik
Abstract Kaczorowski, Mr Seth Cheetham, Ms Franziska
It has long been established that innate Gruhl, Associate Professor Marcel Dinger
immune cells elicit a broad inflammatory
program as a rapid first-line of defence Complex coding and noncoding
transcriptional dynamics in early stem cell
against pathogens. Less well understood are
development
pathogen-specific responses, which vary by
the type of inflammatory molecule, amplitude Abstract
and/or kinetics of inflammatory response. The Low cost-high throughput sequencing
ability of host innate immune cells to direct technology is allowing researchers to revisit
specialist inflammatory responses has now questions previously too costly or complex to
been described for numerous pathogens - answer. One example of this is transcriptional
across many taxonomic levels, right down to profiling studies of development, which are
individual species and strains of pathogen. typically sampled at only a small number of
As the spectrum of innate cell phenotypes time points. In the case of complex mammalian
produced by these inflammatory programs embryonic stem (ES) cell development, such
drives long term adaptive immunity, it is profiling is likely misses key developmental
important that the factors underlying their transitions. Furthermore, as RNA signals and
diversity are understood. Current pathway or especially noncoding RNAs can be extremely
network models of innate immune signalling labile, it is probable that our understanding
do not reflect phenotypic diversity, or of transcriptional events underlying
provide a mechanistic explanation for the normal development –and consequently
evolution of divergent molecular responses developmental disease- is missing significant
from convergent signalling pathways. While detail. We have performed RNA sequencing
this convergence is driven from evolutionary in unprecedented temporal detail on the first
conservation of the genome, transcriptomic 5 days of mouse ES cell development and are
diversification mechanisms may provide a unraveling exquisite complexity of coding and
means to contribute to diversity in innate noncoding gene dynamics.
immune responses. This temporal detail has allowed us to observe
We have used Cap Analysis Gene Expression previously unknown variation in expression
(CAGE) to characterise common and unique of well-characterised developmental genes
components of monocyte transcriptional as well as significantly increased regulatory
programs elicited by representative pathogens. complexity. We have identified hundreds
Within these programs we have identified of rapidly cycling (<24hr cycle) coding and
candidate genes subject to regulation by non-coding transcripts in ES development.
multiple transcription start sites (TSSs). Our Additionally. We have also identified
findings suggest that alternate TSSs provide transcriptional networks of protein coding
115
Handbook
2014 AGTA
Conference
genes (enriched for cell cycle, focal adhesion gene containing that exon under a certain
and transcription factors) responding to comparison. Similarly to the gene level analysis
expression changes of novel noncoding RNAs. of RNA-Seq data, we use negative binomial
Finally, this data is enabling unprecedented distribution to model exon counts. An empirical
capacity to unravel driver/carrier Bayes information sharing strategy is applied
transcriptional dynamics at complex gene loci to estimate the variation between biological
(bi-directional/antisense). replicates. Statistical tests are performed at
Fine temporal resolution has highlighted both the exon level and the gene level. The
important gene expression changes in early method has been implemented in the edgeR
differentiation and identified many noncoding package.
transcripts with a potential driver role in cellular
development. It is becoming possible to
elucidate the complexities of transcriptional Poster 91
control during cellular differentiation.
Dr Wei Shi, Dr Bhupinder Pal, Prof. Geoffrey

Lindeman, Prof. Jane Visvader, Prof. Gordon
Poster 90
Smyth
Characterizing changes in epigenome and

Dr Yunshun Chen, Professor Gordon Smyth transcriptome during mammary stem cell
differentiation
Detecting Differential Exon Usage
in RNA-Seq Abstract
Understanding the molecular mechanisms
Abstract
regulating mammary stem cell differentiation
Alternative splicing is a process where exons
is not only important for elucidating this
are differentially combined or skipped,
important biological process, but also
resulting in multiple protein isoforms encoded
important for discovering genes that may
by a single gene. It generates diverse
contribute to the initiation and progression of
transcripts and provides an opportunity for
diseases relevant to this process such as breast
gene regulation. The proteins translated
cancer.
from alternatively spliced mRNAs will contain
different or even opposite biological functions. We have performed a next-generation
It has been discovered that more than 95% sequencing experiment to profile the global
of human multi-exon genes express multiple changes in histone marks including H3K4me3
splice isoforms. Alternative splicing can and H3K27me, when mammary stem cells
affect various functions in cellular processes, differentiate into luminal progenitors and
tissue specificity, developmental states and mature luminal cells. We have also carried
disease conditions. Recent high throughput out a whole genome expression microarray
technologies, such as RNA-Seq, have been experiment to measure the gene expression
proven to be very powerful tools in studying changes during this differentiation process.
alternative splicing. Meanwhile, it is also very We have developed a novel bioinformatics
tempting and challenging for statisticians method to characterize the changes of histone
and bioinformaticians to develop statistical marks genes and to correlate them with global
methods for detecting alternatively splicing gene expression changes.
events. Here we present a method for testing Results: We found that both H3K4me3 and
for differential exon usage between different H3K27me3 played important roles in regulating
experimental conditions or different genetic mammary stem differentiation. H3K4me3
backgrounds. In particular, we compare the was found to induce gene expression during
change of the expression level of each exon differentiation, ie. genes that were turned on
to the change of the expression level of the acquired higher level of H3K4me3 marks and
116
Handbook
2014 AGTA
Conference
genes that were turned off had reduced level

Poster 92
of H3K4me3. However, H3K27me3 was found
to have a opposite role, ie. repressing gene
expression during differentiation.
Dr Susan Corley, Mr Cesar Canales, Professor
This study shed new lights on how the Edna Hardeman, Dr Stephen Palmer, Professor
differentiation of mammary stem cells Marc Wilkins
were regulated at the molecular level. The
RNA-Seq analysis of lip tissue from a mouse
bioinformatics analysis strategy we developed
model of Williams-Beuren Syndrome
in this study has been demonstrated to be very
successful in charactering global changes in Abstract
the epigenome and transcriptome and in the Williams-Beuren Syndrome (WBS) is a genetic
correlations between them. disorder associated with multisystemic
abnormalities, including craniofacial
dysmorphology and cognitive defects. It is
caused by a hemizygous microdeletion involving
up to 28 genes in chromosome 7q11.23.
Genotype/phenotype analysis of atypical
microdeletions implicates two evolutionary-
related transcription factors, GTF2I and
GTF2IRD1, as prime candidates for the cause
of the facial dysmorphology. Using a targeted
Gtf2ird1 knockout mouse, we have employed
massively-parallel sequencing of mRNA (RNA-
Seq) to understand changes in the transcriptional
landscape associated with absence of Gtf2ird1.
This in turn provides insight into the proteins
likely to be activated or repressed by this
transcriptional regulator. RNA-Seq data (100 bp
paired-end reads) from lip tissue of 3 KO mice
and 3 controls were mapped to the mouse
genome (mm10) with TopHat2 and differential
expression analysis was performed using edgeR
and DESeq2. Our results show that the absence of
GTF2IRD1 is associated with increased expression
of genes involved in cellular proliferation,
including growth factors, consistent with the
observed phenotype of extreme thickening of
the epidermal layer. We also found a decrease
in expression of genes involved in certain
signalling pathways including Wnt signalling,
indicating dysregulation in the complex
signalling network necessary for epidermal
differentiation and facial skin patterning.
117
Handbook
2014 AGTA
Conference
Poster 94
Poster 93
Dr Sonika Tyagi
Amanda Chamberlain, Christy Vander Jagt,
Michael Goddard, Leah Marrett, Ben Hayes Stranded versus non-stranded RNA-seq
Discovery of allele specific expression in Abstract

RNAseq data from eighteen bovine tissues RNA-seq has proven to be a powerful
technique for transcriptome analysis and has
Abstract
changed the way in which transcription can be
A common finding in genome wide association
viewed. RNA-seq can capture both sequence
studies in humans, cattle and other species is
and abundance of the RNA species in a
that many significant hits fall outside genes. One
transcriptome, which makes it an attractive
possibility is that regulatory variants underpin
approach to study structure of the transcripts,
a large proportion of complex trait variation.
splice variants and transcriptional activities in
Detecting allele specific expression (ASE) is
non-coding, intergenic or untraslated regions.
one potential method for finding regulatory
Following recent proliferation of RNA-seq
variants. In order to investigate the frequency,
sample preparation protocols, there is little
and consistency across tissues, of ASE in cattle,
known about effect of various library protocols
we sequenced the RNA of 18 tissues from a
on the downstream analysis is incomplete.
lactating cow, collected at a single point in time.
To correct for mapping bias, reads were mapped In this paper we will discuss some of the
to parental genomes and then a chi squared common library preparation biases their effects
test was performed to test for ASE at known on downstream analysis. We will be specifically
heterozygous SNP within exons and therefore comparing stranded and unstranded RNA-seq
genes. Results confirmed 4 genes to be imprinted library preparation methods with respect to
in cattle and discovered 13 genes previously whole transcriptome analysis.
found to be imprinted in mice or humans to
be imprinted in cattle. After removing these
imprinted genes 41,805 SNP were tested within
11,568 genes for ASE. 68% of these genes showed
significant ASE in at least one tissue, within
any one of the tissues 9-43% of genes showed
significant ASE. The major allele expressed varied
from ~55% to 100% of transcripts. 84% of the
genes showing ASE did so in a subset of tissues.
In general, the results suggest there are many
genes in cattle that could be under cis regulatory
control. This dataset provides information that
will be used to find causative variants in these
regulatory regions. Such causative variants
would then be used to improve the accuracy
of genomic selection in cattle.
118
Handbook
2014 AGTA
Conference
THANK YOU The organising committee for the 2014

AGTA Conference would like to take
TO OUR GENEROUS
the opportunity to thank the following
organisations for their valuable contribution,
and ask that all delegates please consider
SPONSORS these organisations for future business
relationships.
Gold Sponsors
Illumina is advancing human health by Pacific Biosciences has developed a

unlocking the power of the genome; from novel approach to studying the synthesis
genome-wide discovery through to targeted and regulation of DNA, RNA and
validation and beyond. proteins. Combining recent advances in
At Illumina, our goal is to apply innovative nanofabrication, biochemistry, molecular
technologies and revolutionary assays to the biology, surface chemistry and optics,
analysis of genetic variation and function, Pacific Biosciences has created a powerful
making studies possible that were not even technology platform called single molecule,
imaginable just a few years ago. real-time, or SMRT®, technology.
These revolutionary tools for DNA, RNA and SMRT technology enables real-time analysis of
protein analysis are enabling rapid advances in biomolecules with single molecule resolution,
disease research, drug development and the which has the potential to transform the
development of molecular tests in the clinic. understanding of biological systems by
providing a window into these systems that has
www.illumina.com
not previously been open for scientific study.
www.pacificbiosciences.com
Millennium Science is a specialist distribution Fluidigm develops and manufactures

company established in 1999 specifically to Integrated Fluidic Circuit (IFC) technology
source, supply and support high technology that significantly improves productivity in life
solutions to the Australasian Life Science science research. Fluidigm IFCs enable the
market. The company is well established simultaneous performance of thousands of
as a leading distributor of instrumentation sophisticated biochemical measurements in
and reagents for the research market with nanolitre volumes. These systems are used for
recognition of performance and business applications ranging from single-cell genomics,
practice excellence by the global companies targeted resequencing, gene expression and
represented, including Fluidigm and Pacific digital PCR.
Biosciences.
www.fluidigm.com
www.mscience.com.au
119
Handbook
2014 AGTA
Conference
Bronze Sponsor Conference Supporters
Genome Biology publishes research articles,

Bio-Strategy is a team of highly qualified new methods and software tools, in addition to
people, throughout Australia and New reviews and opinions, from the full spectrum of
Zealand, who share a passion for Science. biology, including molecular, cellular, organism
Representing world leading suppliers, Bio- or population biology studied from a genomic
Strategy brings our customers a complete perspective, as well as sequence analysis,
range of solutions, from basic laboratory bioinformatics, proteomics, comparative
instrumentation through to advanced biology and evolution.
technology, covering applications in the www.genomebiology.com
Research, Healthcare, Biotechnology and
Industrial Markets.
Our focus and commitment to our customers
and suppliers, is key to Bio-Strategy’s
excellence and success.
www.petermac.org
www.bio-strategy.com
www.qfab.org
www.combine.org.au
120
Handbook
2014 AGTA
Conference
THANK YOU TO OUR

TRADE EXHIBITORS
Trade Booth 1 Trade Booth 3
Illumina is advancing human health by unlocking Thermo Fisher Scientific is the world leader in
the power of the genome; from genome-wide serving science, enabling our customers to
discovery through to targeted validation and make the world healthier, cleaner and safer.
beyond. We are the leading provider of analytical
At Illumina, our goal is to apply innovative instruments, equipment, reagents and
technologies and revolutionary assays to the consumables, software and services for
analysis of genetic variation and function, research, analysis, discovery and diagnostics.
making studies possible that were not even In Australia and New Zealand, you can
imaginable just a few years ago. access the renowned Thermo Scientific
portfolio as well as premium, through Thermo
These revolutionary tools for DNA, RNA and
Fisher Scientific. Our comprehensive range
protein analysis are enabling rapid advances in
includes high-end analytical and process
disease research, drug development and the
instrumentation, laboratory equipment, a
development of molecular tests in the clinic.
complete range of consumables, chemicals
www.illumina.com and reagents as well as the service and
support to optimise your business.
www.thermofisher.com.au
Trade Booth 2
Trade Booth 4
Australian Genome Research Facility Ltd (AGRF),

a not for profit company, is Australia’s largest
provider of genomics services and solutions. AGRF
Life Technologies™ products harness the
has laboratories in Brisbane, Sydney, Melbourne,
power of science to transform lives. As a
Adelaide and Perth, each providing a gateway
member of the Thermo Fisher Scientific family
to a national network of state-of-the-art facilities,
of brands, our instruments, everyday tools,
technology and expertise. From microbes to plants,
and services offer high-quality, innovative
animals and humans, AGRF provides a full range
life science solutions for every lab and
of services across the entire biological spectrum.
application.
We provide services to academia and industry with
clients from bioscience, environmental science, www.lifetechnologies.com
biomedicine and agricultural biotechnology.
Our range of services is unique in Australia and
recognised internationally. AGRF is accredited by
NATA to the ISO/IEC17025:2005 Quality Standard.
www.agrf.org.au
121
Handbook
2014 AGTA
Conference
Tecan is a leading global provider of Agilent is a global leader in target enrichment

laboratory instruments and solutions in solutions for next-generation sequencing and
biopharmaceuticals, forensics and clinical genomic microarrays. By synthesizing custom
diagnostics. We specialize in the development, complex mixtures of long oligonucleotides,
production and distribution of automated SureSelect and HaloPlex target enrichment
workflow solutions for laboratories in the life solutions enable researchers to identify
sciences sector. genomic regions of interest for focused, cost-
We offer expandable, reconfigurable solutions effective variant profiling, with workflows that
that cover all your genomic automation needs generate sequencing-ready libraries in just
including DNA extraction, Normalization, one day. For more information visit Agilent
PCR set up, DNA shearing and NGS sample Genomics.
preparation. For example, assay-specific www.genomics.agilent.com
set up for PCR, SNP scoring, sequencing,
reaction clean up, as well as library fragment
preparation and emPCR enrichment for next Trade Booth 8 & 9
generation on the same platform makes the
Tecan Freedom EVO a truly powerful tool for
the genomics laboratory.
www.tecan.com
Millennium Science is a specialist distribution
company established in 1999 specifically to
source, supply and support high technology
Trade Booth 6 solutions to the Australasian Life Science
market. The company is well established
as a leading distributor of instrumentation
and reagents for the research market with
Roche offers research solutions, from nucleic recognition of performance and business
acid extraction and qPCR through to targeted practice excellence by the global companies
enrichment using NimbleGen SeqCap represented, including Fluidigm and Pacific
reagents. Biosciences.
Roche is committed long-term to sequencing, Pacific Biosciences has developed a

recently acquiring Genia Technologies and novel approach to studying the synthesis
investing in Stratos Genomics, both for single and regulation of DNA, RNA and
molecule sequencing nanopore technologies; proteins. Combining recent advances in
the goal being higher throughput, short nanofabrication, biochemistry, molecular
turnaround and reduced costs. biology, surface chemistry and optics,
Pacific Biosciences has created a powerful
www.lifescience.roche.com
technology platform called single molecule,
real-time, or SMRT®, technology.
122
Handbook
2014 AGTA
Conference
SMRT technology enables real-time analysis of Trade Booth 11

biomolecules with single molecule resolution,
which has the potential to transform the
understanding of biological systems by
providing a window into these systems that has
not previously been open for scientific study.
Fluidigm develops and manufactures
The Ramaciotti Centre is a leading genomic
Integrated Fluidic Circuit (IFC) technology
service provider located at the University
that significantly improves productivity in life
of New South Wales, Sydney. Our services
science research. Fluidigm IFCs enable the
include next-generation sequencing, Sanger
simultaneous performance of thousands of
sequencing, microarray, SNP genotyping, CNV
sophisticated biochemical measurements in
analysis and gene expression analysis including
nanolitre volumes. These systems are used for
single cell gene expression studies.
applications ranging from single-cell genomics,
targeted resequencing, gene expression and We produce data of the highest quality and
digital PCR provide excellent personalised customer
service. Our professional team has many years
www.mscience.com.au
of experience facilitating projects from design
www.fluidigm.com through to downstream analysis.
www.pacificbiosciences.com We are a CSPro certified provider of Illumina
sequencing and an Authorised Affymetrix
Service Provider.
Trade Booth 10 www.ramaciotti.unsw.edu.au
Trade Booth 12
To be your premier scientific partner, BGI Tech
is committed to providing high quality and
innovative products and solutions in biological,
agricultural and medical research.
Genesearch is the Australian distributor of
BGI Tech was founded in 2012 and is New England Biolabs products, including
headquartered in Shenzhen, China. Our service the NEBNext product range for NGS library
network covers more than 100 countries and preparation. NEBNext reagents provide
regions. Currently, BGI Tech collaborates with solutions for the major NGS platforms. Featuring
over 10,000 organizations and 30,000 partners, NEB’s NGS-optimized polymerase, Q5, for robust
providing highly reliable services for thousands performance with a wide range of amplicons,
of research projects, including several large- regardless of GC content. Please visit the NEB
scale international genome projects. To website to see the latest products including 96
date, BGI Tech has contributed to over 600 Dual Index Primers compatible with the Illumina
publications, with more than 100 articles platform.
appearing in top-tier journals such as Nature
www.genesearch.com.au
and Science.
www.bgitechsolutions.com
123
Handbook
2014 AGTA
Conference
Bio-Strategy is a team of highly qualified Sarstedt Australia is recognised in Australia

people, throughout Australia and New and internationally as a quality manufacturer
Zealand, who share a passion for Science. of a comprehensive range of diagnostic,
Representing world leading suppliers, Bio- laboratory and medical products that includes
Strategy brings our customers a complete both equipment and consumables for use in
range of solutions, from basic laboratory pathology, laboratory, research, transfusion
instrumentation through to advanced medicine, blood banking and hospital/
technology, covering applications in the incontinence applications. Sarstedt’s product
Research, Healthcare, Biotechnology and range includes plastic disposables for the
Industrial Markets. sophisticated requirements of cell culture
research and PCR applications. The range
Our focus and commitment to our customers
includes Serological Pipettes, Tissue Culture
and suppliers, is key to Bio-Strategy’s
Flasks, TC Dishes and TC Multi Well Plates, Filters,
excellence and success.
PCR Plates, PCR Strips and Cups, Adhesive
www.bio-strategy.com Seals, Cryo Preservation Tubes, as well as a
comprehensive range of general purpose
laboratory tubes and plastic consumables.
Trade Booth 14 www.sarstedt.com
Trade Booth 16
QIAGEN is a global leader in sample and assay

technologies for life sciences, applied testing
and molecular diagnostics. Its products are
TrendBio is a leading supplier of specialised
considered standards that comprise complete
instrumentation for the Life Sciences. During
solutions from sample to result. QIAGEN offers
the AGTA 2014 meeting we will be promoting;
the broadest portfolio of molecular diagnostic
assays for infectious diseases including the only • DNA Shearing (Covaris)
test for human papillomavirus (HPV), which has • FFPE Extraction (DNA/RNA)
both FDA and CE-approvals. The company • Chromatin Shearing
has developed a comprehensive portfolio • RNA and DNA Analysis (capillary gel
of more than 500 proprietary, consumable electrophoresis)
products and instruments for sample collection, • Microscale Thermophoresis – Biomolecule
nucleic acid and protein handling, separation, Interactions (NanoTemper)
purification, detection, and open and target • Automated Western Blotting (Protein
specific assays. Simple)
• Pre-Clinical Imaging
www.qiagen.com
www.trendbio.com.au
124
Handbook
2014 AGTA
Conference
Integrated DNA Technologies (IDT) is a

The MHTP Medical Genomics Facility at the
leader in manufacturing and developing
Monash Health Translation Precinct (MHTP) in
products for the research and diagnostic life
Victoria uses the latest and most innovative
science market. IDT is also the world leader
genomic technologies to service research
in nucleic acid synthesis and serves the areas
institutes throughout Australia. We maintain an
of academic research, biotechnology, and
excellent reputation for provision of the highest
pharmaceutical development.
quality genomic data and client support.
Product offerings include DNA and RNA
Our not-for-profit, NATA-accredited
oligos, qPCR assays and probes, siRNA
facility offers Sanger and Next Generation
duplexes, DNA fragments and custom
Sequencing, High Content Screening,
gene synthesis. Pioneering methods in oligo
Microarray, Real-Time qPCR and Fluidigm
synthesis and quality control allow IDT to be
Single Cell technologies. Combined with our
able to synthesize DNA oligos up to 2kb. IDT
professional consultation, we aim to facilitate
products support applications such as DNA
and accelerate genomic-based translational
sequencing, DNA amplification, SNP detection,
research.
microarray analysis, expression profiling, gene
www.mimr-phi.org quantification, and functional genomics.
www.idtdna.com
Trade Booth 18
On Q
software
OnQ Software is dedicated in providing
leading edge custom technology solutions with
the aim of enabling end to end integration
across your entire workplace using our suite
of state of the art software, tools and highly
trained experts who can tailor a solution for
your business needs.
www.onqsoft.com.au
125
Handbook
2014 AGTA
Conference
Scientifix is an Australian-owned distribution Eppendorf is a leading life science company

company established in 1998. Our product that develops and sells instruments,
lines include Macherey Nagel German-quality consumables, and services for Liquid, Sample
extraction kits for all your DNA/RNA and and Cell handling in laboratories worldwide.
clean-up purification requirements. Macherey Eppendorf’s product range includes pipettes
Nagel isolation products are also formatted and automated pipetting systems, dispensers,
and optimised for seamless automation centrifuges, mixers, spectrometers and DNA
on a number of leading robotic platforms. amplification equipment as well as ultra-low
Macherey Nagel also offers magnetic particles temperature freezers, platform shakers, CO2
for reaction clean-up and size selection in NGS Incubators, Bioprocess equipment and cell
library construction. manipulation systems. Associated consumables
Clontech Laboratories is a leading supplier like pipette tips, test tubes, microtiter plates,
for innovative kits and reagents for gene cell culture consumables and disposable
discovery, regulation, and function. Clontech’s bioreactors vessels complement the instruments
SMARTer cDNA synthesis is the industry standard for the highest quality workflow solutions.
for exceptional sensitivity and complete Our mission is to be a synonym for customer-
transcriptome coverage as it synthesizes focused processes, innovative technologies,
accurate, full-length cDNA directly from cells or premium products and services to improve
total RNA. SMARTer Ultra Low Input RNA Kit for human living conditions.
Sequencing – v3 is the gold standard in cDNA
www.eppendorf.com.au
Synthesis for Single-Cell Transcriptome Analysis
and compatible with leading NGS platforms.
Takara Bio has positioned itself as a leading
Trade Booth 22
manufacturing company in the Japanese life
science market. Takara Bio’s R&D scientists
are expert enzymologists who are continually
working to improve enzymes and buffers for SI Systems provides IT services to Life Science
PCR, real-time PCR, and RT-PCR. Markets and specialises in enterprise IT
www.scientifix.com.au Infrastructure solutions for Big data storage,
custom software development and
Bioinformatics data analysis services. We
can help you choose the right data storage
solutions ranging from 10TB to hundreds of PB
(EMC: Isilon,VNX, Pivoltal or Coraid: Flash, Deep
storage), along with CISCO and Dell compute
systems. Contact us for expert Bioinformatic
solutions for analysis and biological
interpretation of high-throughput sequencing
data.
www.sisystems.com.au
126
Handbook
2014 AGTA
Conference
Trade Both 23 Trade Both 24
DNA Genotek provides high-quality biological

PerkinElmer is a premier provider of cutting- sample collection, stabilization and preparation
edge technologies and solutions enabling products for human genetics, microbiology
researchers to create more efficient and and animal genetics. The company’s products
quicker life-saving and enhancing medicines protect and stabilize multiple sample types for
and diagnostic tests. As a leading solutions long-term storage at ambient temperature to
innovator, PerkinElmer bridges the gap ensure the highest quality results for genetic
between in vitro assays and in vivo results, analysis and testing. Due to their reliability and
allowing researchers to translate their results ease-of-use, DNA Genotek’s products are used
into cures for human disease. by thousands of academic, biotechnology,
Our wide range of systems consist of diagnostic, agriculture, and other leading
instrumentation for automation and detection institutions around the globe.
solutions, cellular imaging and analysis www.dnagenotek.com
hardware and software, and an industry-
leading portfolio of consumables products. Our
drug discovery and research reagents support
Trade Booth 25
our core expertise in cellular sciences, time-
resolved fluorescence, chemiluminescence,
radioactive labelling, and the detection of
proteins and nucleic acids.
The company’s offerings also include state- Bio-Rad Laboratories, Inc. Designs,
of-the-art microfluidics, lab automation & manufactures, and distributes a broad
liquid handling, tissue microscopy, preclinical range of innovative products and solutions
imaging technologies, and discovery & for the life science research and clinical
development outsourcing solution. diagnostic markets. Bio-Rad is renowned for its
commitment to quality and customer service
www.perkinelmer.com
among university and research institutions,
hospitals, public health and commercial
laboratories, as well as the biotechnology,
pharmaceutical, and food safety industries.
Founded in 1952, Bio-Rad is based in California,
and serves more than 100,000 research and
healthcare industry customers globally.
www.bio-rad.com
127
Handbook
2014 AGTA
Conference
Macrogen is the current market leader in

biotechnology services based in Korea,
providing pivotal research tools to a global GeneWorks is an Adelaide-based genomic
client base. The company provides high services provider and supplier of innovative
quality genomic sequencing, microarray, oligo products for molecular research. Supported
synthesis, knock-out mouse, and bioinformatics technologies include MassARRAY from Agena
services at competitive prices. Using only the Bioscience, PCR reagents and NGS kits from
most up to date technology available on the KAPA Biosystems, SAGE Science Pippin Prep
market, Macrogen is able to facilitate large and Blue Pippin for library preparation, Exiqon
scale evolutionary studies, whilst still taking miRNA solutions, and KASP genotyping from
care to provide smaller scale investigations, LGC. Genetic analysis services include SNP
customised to the client’s needs. genotyping assays (custom designed and
www.macrogen.com Agena panels). GeneWorks now also offers a
sample validation service using Agena’s Exome
ID panel.
www.geneworks.com.au
128
Handbook
2014 AGTA
Conference
EXHIBITION
FLOOR PLAN
EXHIBITOR BOOTH
Illumina 1
AGRF 2
Thermo Fisher 3
Scientific
Life Technologies 4
Tecan Australia 5
Roche 6
Agilent Technologies 7
Millennium Science 8&9
BGI Tech 10
Ramaciotti Centre 11
for Genomics
Genesearch 12
Bio-Strategy 13
AGTA Qiagen 14
REGISTRATION
Sarstedt 15
TrendBio 16
MIMR - PHI Institute 17
17 18 (Monash Health)
OnQ Software 18
16 19
Intergrated DNA 19
Scientifix LIFE 20
15 20 Eppendorf 21
14 21 SI Systems 22
PerkinElmer 23
DNA Genotek 24
13
Bio-Rad 25
12 22 Macrogen 26
GeneWorks 27
25 24 23
26
27 1
11
AGTA
2 SPEAKER
10
AGTA PREPARATION
PLENARY 9 3
8 4
AGTA
7 6 5 BREAKOUT
129
Handbook
2014 AGTA
Conference
AGTA 2014
DELEGATE LIST
First Name Last Name Organisation State Country

Matias Abregu PerkinElmer VIC Australia
Satoru Akama Tokyo Institute Of Technology Japan
Raymond Ali University Of Tasmania TAS Australia
Richard Allcock University Of Western Australia WA Australia
Janelle Allen Illumina Australia Pty Ltd VIC Australia
Mohammed Alnakhli Flinders University SA Australia
Mohammed Alqahtani CEGMR Saudi Arabia
Mary Anderson SydPath, St Vincent’s Hospital NSW Australia
Stuart Archer Monash University VIC Australia
Josh Ardley Tecan Australia VIC Australia
Nicola Armstrong University Of Sydney NSW Australia
Marie-Liesse Asselin-Labat The Walter & Eliza Hall Institute Of VIC Australia
Medical Research
Kelly Atkinson University Of Auckland New Zealand
Melanie Bahlo The Walter & Eliza Hall Institute Of VIC Australia
Medical Research
Gia Bawa Geneworks Pty Ltd SA Australia
Anthony Beckhouse Bio-Rad Labratories NSW Australia
Alex Beesley Telethon Kids Institute WA Australia
Katrina Bell Murdoch Childrens Research Institute VIC Australia
Nigel Bennett University Of Queensland QLD Australia
Marnie Blewitt The Walter & Eliza Hall Institute Of VIC Australia
Medical Research
Justin Borevitz Australian National University ACT Australia
David Bowtell Peter MacCallum Cancer Centre VIC Australia
Titus Brown Department of Microbiology and USA
Molecular Genetics
Magdalena Budzinska Centenary Institute NSW Australia
Suzanne Butcher Australian Institute For Bioengineering QLD Australia
And Nanotechnology
Keith Byron Healthscope VIC Australia
Sophia Cameron- University Of Otago New Zealand
Christie
Peter Campbell Wellcome Trust Sanger Institute United
Kingdom
Minh Duc Cao The University Of Queensland QLD Australia
Karen Carpenter PathWest Diagnostic Genomics WA Australia
Kim Carter Telethon Kids Institute WA Australia
130
Handbook
2014 AGTA
Conference

Daniel Catchpoole Biospecimens Research And Tumour NSW Australia
Bank, Children’s Cancer Research
Unit
Alistair Chalk St Vincents Institute VIC Australia
Amanda Chamberlain Department Of Environment And VIC Australia
Primary Industries
Anandhakumar Chandran Kyoto University Japan
Jac Charlesworth University Of Tasmania TAS Australia
Rima Chaudhuri Garvan Institute Of Medical Research NSW Australia
Yunshun Chen The Walter & Eliza Hall Institute Of VIC Australia
Medical Research
Wenhan Chen University Of Queensland QLD Australia
Tyrone Chen University Of Queensland QLD Australia
Belinda Chong Victorian Clinical Genetics Services VIC Australia
Cheuk Wang Chung National University Of Singapore Singapore
Nicole Cloonan QIMR Berghofer Medical Research QLD Australia
Institute
Ana Conesa Genomics of Gene Expression Lab Spain
Thomas Conway IBM Research Australia VIC Australia
Vincent Corbin The Walter & Eliza Hall Institute Of VIC Australia
Medical Research
Susan Corley SBI, University Of NSW NSW Australia
Mark Cowley Garvan Institute Of Medical Research NSW Australia
Jeffrey Craig Murdoch Childrens Research Institute VIC Australia
Mark Crowe QFAB Bioinformatics QLD Australia
Tamsyn Crowley Deakin University VIC Australia
Mark Cruickshank Telethon Kids Institute WA Australia
Helen Cumming Mimr-Phi VIC Australia
Robert Curtain Bio-Rad Laboratories NSW Australia
Patrick Danoy Roche Diagnostics NSW Australia
Aaron Darling University Of Technology Sydney NSW Australia
Harriet Dashnow Murdoch Childrens Research Institute VIC Australia
Nadia Davidson Murdoch Childrens Research Institute VIC Australia
Mark Davis PathWest Diagnostic Genomics WA Australia
Sarah-Jane Dawson Peter MacCallum Cancer Centre VIC Australia
Niantao Deng Garvan Institute Of Medical Research NSW Australia
Leanne Dibbens Sansom Institute For Health Research, SA Australia
School Of Pharmacy And Medical
Sciences, U
Alison Digney Life Technologies VIC Australia
Marcel Dinger Garvan Institute Of Medical Research NSW Australia
James Doecke CSIRO QLD Australia
Kenneth Doig Peter MacCallum Cancer Centre VIC Australia
Maria Doyle Peter MacCallum Cancer Centre VIC Australia
131
Handbook
2014 AGTA
Conference

Desiree Du Sart Victorian Clinical Genetics Services VIC Australia
Bjorn Dueholm University Of South Australia SA Australia
Ankit Dutta The University Of Adelaide SA Australia
Jason Ellul Peter MacCallum Cancer Centre VIC Australia
Michael Erlichster The Walter & Eliza Hall Institute Of VIC Australia
Medical Research
Kelly Ewen-White Life Technologies VIC Australia
Miriam Fanjul Fernandez Murdoch Childrens Research Institute VIC Australia
Cassy Faux Australian Antarctic Division TAS Australia
Andrew Fellowes Peter MacCallum Cancer Centre VIC Australia
Christoffer Flensburg McRi VIC Australia
Cathy Ford DNA Genotek Canada
Sue Forrest Australian Genome Research Facility VIC Australia
Archa Fox The Harry Perkins Institute of Medical WA Australia
Research, UWA
Iain Fraser National Institute of Health USA
Michelle French AgResearch Ltd New Zealand
Saskia Freytag The Walter & Eliza Hall Institute Of VIC Australia
Medical Research
Melissa Fullwood Cancer Science Institute Singapore Singapore
Devika Ganesamoorthy Institute For Molecular Biosciences QLD Australia
Jonathan Gannoulis OnQ Software VIC Australia
Amee George Peter MacCallum Cancer Centre VIC Australia
Brian Gloss Kinghorn Cancer Centre NSW Australia
Anita Goldinger University Of Queensland QLD Australia
David Goode Peter MacCallum Cancer Centre VIC Australia
Jodee Gould MHTP, Medical Genomics Facility VIC Australia
Alexander Gout Telethon Kids Institute WA Australia
Parry Guilford University of Otago New Zealand
Richard Harrison Bio-Rad Laboratories NSW Australia
Julie Haste Agilent Technologies VIC Australia
Jean Hatinguais Kapa Biosystems USA
Christophe
Lyndal Henden The Walter & Eliza Hall Institute Of VIC Australia
Medical Research
Duncan Hewett University Of Adelaide SA Australia
Sarah Hickey Life Technologies VIC Australia
Tony Hoady Eppendorf NSW Australia
Aliaksei Holik The Walter & Eliza Hall Institute Of VIC Australia
Medical Research
Kathryn Holt University of Melbourne VIC Australia
Katharine Howell ARC CoE In Plant Energy Biology, WA Australia
University Of Western Australia
132
Handbook
2014 AGTA
Conference

Michelle Howie Roche Diagnostics NSW Australia
Ji-Fan (Sarah) Hsieh The Australian National University ACT Australia
Siobhan Hughes King’s College London United
Kingdom
Sally Hunter Peter MacCallum Cancer Centre VIC Australia
Gyorgy Hutvagner UTS NSW Australia
Michael Inouye University Of Melbourne VIC Australia
Kaushala Jayawardana University Of Sydney NSW Australia
Aaron Jeffs University Of Otago New Zealand
Margaret Jordan Comparative Genomics Centre QLD Australia
Chol-Hee Jung VLSCI Life Sciences Computation VIC Australia
Centre
David Kainer Australian National University ACT Australia
Naga Kasinadhuni AGRF QLD Australia
Gurpreet Kaur Millennium Science VIC Australia
Anabel Kearney Liverpool Hospital NSW Australia
Farzaneh Kordbacheh University Of Queensland QLD Australia
Carsten Kulheim Australian National University ACT Australia
Paul Lacaze Millennium Science VIC Australia
Nicole Lake Murdoch Childrens Research Institute VIC Australia
Timo Lassmann Telethon Kids Institute WA Australia
Charity Law University Of Zurich Switzerland
David Lawrence ACRF Cancer Genomics Facility, SA Australia
Centre For Cancer Biology
Tim Lawrence NZGL/Auckland University New Zealand
Lien Le Bioinformatics Resource Australia Of QLD Australia
EMBL Australia (BRAEMBL)
Marcus Lefebure Peter MacCallum Cancer Centre VIC Australia
Xuan Li BGI Tech Solutions Co. Ltd China
Xi Li CSIRO ACT Australia
Jason Li Peter MacCallum Cancer Centre VIC Australia
Webber W.P. Liao Bioinformatics Resource EMBL QLD Australia
Australia; NCI Specialised Facility In
Bioinformati
Ruby Lin University Of New South Wales NSW Australia
Tao Liu Children’s Cancer Institute Australia NSW Australia
Jianhua Liu Institute For Comprehensive Utilisation China
Of Marine Biological Resources
Sheng Liu The Walter & Eliza Hall Institute Of VIC Australia
Medical Research
Andrew Lonsdale The University Of Melbourne VIC Australia
Evangelina López De Spanish National Cancer Research Spain
Maturana Centre
133
Handbook
2014 AGTA
Conference

David Lovell CSIRO/Australian Bioinformatics ACT Australia
Network
Robyn Lukeis SydPath, St. Vincent’s Hospital NSW Australia
Vesna Lukic The Walter & Eliza Hall Institute Of VIC Australia
Medical Research
Aaron Lun The Walter & Eliza Hall Institute Of VIC Australia
Medical Research
Richard Lupat Peter MacCallum Cancer Centre VIC Australia
Donia Macartney- Institute Of Environmental Science New Zealand
Coxson And Research
Ruth MacKinnon Victorian Cancer Cytogenetics VIC Australia
Service
Paul Maclean AgResearch Ltd New Zealand
Gregory Maes James Cook University QLD Australia
Ian Majewski The Walter & Eliza Hall Institute Of VIC Australia
Medical Research
Jovana Maksimovic Murdoch Childrens Research Institute VIC Australia
Nauman Maqbool AgResearch Ltd New Zealand
Vikki Marshall Australasian Genomics Technologies QLD Australia
Association
Justine Marum SA Pathology SA Australia
Elizabeth Mason University Of Queensland QLD Australia
Karen Mather University Of New South Wales NSW Australia
John Mattick Garvan Institute of Medical Research NSW Australia
Damara McAndrew Genesearch QLD Australia
Mark McCabe Garvan Institute Of Medical Research NSW Australia
Amy McCart Reed UQCCR QLD Australia
Ann McCormack Garvan Institute Of Medical Research NSW Australia
Helen McCormick Victor Chang Cardiac Research NSW Australia
Institute
George McGillivray Murdoch Childrens Research Institute VIC Australia
Liam McIntyre SA Pathology SA Australia
Cliff Meldrum Hunter New England Health NSW Australia
Rebecca Miller Integrated DNA Technologies Pte NSW Australia
Gisela Mir Peter MacCallum Cancer Centre VIC Australia
Mahtab Mirmomeni IBM Research Australia VIC Australia
Mika Mitropoulos Thermo Fisher Scientific VIC Australia
Cath Moore QIAGEN VIC Australia
Ryan Morin Simon Fraser University Canada
Hayley Mountford Murdoch Childrens Research Institute VIC Australia
Kevin Murray Australian National University ACT Australia
Alexander Myburg University of Pretoria South Africa
Stacey Nelson Bio-Strategy NZ
Kathryn North Murdoch Childrens Research Institute VIC Australia
134
Handbook
2014 AGTA
Conference

Harald Oey La Trobe University VIC Australia
Benjamin Ong Murdoch Childrens Research Institute VIC Australia
Alicia Oshlack Murdoch Childrens Research Institute VIC Australia
Yalchin Oytam CSIRO NSW Australia
Arunkumar Padmanaban Agilent Technologies VIC Australia
Lisa Paiman Healthscope VIC Australia
Tony Papenfuss The Walter & Eliza Hall Institute Of VIC Australia
Medical Research
Dhru Patel OnQ Software VIC Australia
Ralph Patrick University Of Queensland QLD Australia
Ellis Patrick University Of Sydney NSW Australia
John Pearson QIMR Berghofer Medical Research QLD Australia
Institute
Stephen Pederson University Of Adelaide SA Australia
Dilmi Perera Lowy Cancer Research Center, UNSW NSW Australia
Victoria Perreau The Florey VIC Australia
Belinda Phipson Murdoch Childrens Research Institute VIC Australia
Justin Pickering Sarstedt Australia SA Australia
Michael Poidinger Singapore Immunology Network Singapore
Leanne Purins SA Pathology SA Australia
Kelly Quek Queensland Centre For Medical QLD Australia
Genomics
Ashfaqur Rahman CSIRO TAS Australia
Louis Ranjard The University Of Auckland New Zealand
Coralie Reich DEPI VIC Australia
Michael Ricos University Of South Australia SA Australia
Scott Ritchie Medical Systems Biology Lab VIC Australia
Matthew Ritchie The Walter & Eliza Hall Institute Of VIC Australia
Medical Research
Bronwyn Robertson Ramaciotti CentreFor Genomics NSW Australia
Daniel Roden Garvan Institute Of Medical Research NSW Australia
Sandra Romanin Westwick-Farrow Media NSW Australia
Jason Ross CSIRO NSW Australia
Fernando Rossello Victorian Bioinformatics Consortium VIC Australia
Alan Rubin The Walter & Eliza Hall Institute Of VIC Australia
Medical Research
Kalyan Rudraraju Eppendorf NSW Australia
Georgie Ryland Peter MacCallum Cancer Centre VIC Australia
Jafar S. Jabbari Australian Genome Research Facility VIC Australia
Simon Sadedin Murdoch Childrens Research Institute VIC Australia
Mnasri Sameh National Gene Bank Of Tunisia Tunisia
Andreas Schlitzer Singapore Immunology Newtork Singapore
Andreas Schreiber ACRF South Australian Cancer SA Australia
Genomics Facility
135
Handbook
2014 AGTA
Conference

Jan Schroeder The Walter & Eliza Hall Institute Of VIC Australia
Medical Research
Mark Schultz University Of Melbourne VIC Australia
Hamish Scott SA Pathology SA Australia
Robert Sebra ICAHN School of Medicine USA
Torsten Seemann Victorian Bioinformatics Consortium VIC Australia
Tim Semple Peter MacCallum Cancer Centre VIC Australia
Jeremy Shearman National Centre For Genetic Thailand
Engineering And Biotechnology
Wei Shi The Walter & Eliza Hall Institute Of VIC Australia
Medical Research
Christopher Sibley University College London United
Kingdom
Beth Signal Garvan Institute Of Medical Research NSW Australia
Kaylene Simpson Peter MacCallum Cancer Centre VIC Australia
Melanie Smith Victorian Clinical Genetics Services VIC Australia
Adam Smolenski University Of Tasmania TAS Australia
Gordon Smyth The Walter & Eliza Hall Institute Of VIC Australia
Medical Research
Maggie Soliman PerkinElmer VIC Australia
Helen Speirs Ramaciotti CentreFor Genomics NSW Australia
Akanksha Srivastava University Of Western Australia WA Australia
Doris Stangl SA Pathology SA Australia
Jo-Ann Stanton University Of Otago New Zealand
Dorothy Steane University Of Tasmania & University Of TAS Australia
The Sunshine Coast
Phil Stephens Foundation Medicine, Inc. USA
Tim Stuart The University Of Western Australia WA Australia
Siamak Tafavogh University Of Technology Sydney NSW Australia
Rick Tankard The Walter & Eliza Hall Institute Of VIC Australia
Medical Research
Graham Taylor University Of Melbourne VIC Australia
Luisa Teasdale Museum Victoria VIC Australia
Erik Thompson Queensland University of Technology QLD Australia
Daniel Thomson Garvan Institute Of Medical Research NSW Australia
Natalie Thorne The Walter & Eliza Hall Institute Of VIC Australia
Medical Research
Amali Thrimawithana The New Zealand Institute For Plant & New Zealand
Food Research Limited
Peri Tobias University Of Sydney NSW Australia
Kim Ton University Of Otago New Zealand
Richard Tothill Peter MacCallum Cancer Centre VIC Australia
Peter Tsai University Of Auckland New Zealand
Nikki Tsoudis Scientifix VIC Australia
136
Handbook
2014 AGTA
Conference

Sonika Tyagi The Australian Genome Research VIC Australia
Facility
Mark Van Der Hoek SA Pathology SA Australia
Tracey Van Stijn AgResearch Ltd New Zealand
Christy Vander Jagt Department Of Environment And VIC Australia
Primary Industries
Vivien Vasic Monash Health Translation Precinct VIC Australia
Oliver Vasilevski Fluidigm Corporation USA
Violeta Velkoska- Healthscope VIC Australia
Ivanova
Matthew Wakefield Life Sciences Computational Centre, VIC Australia
VLSCI
Mark Waltham St Vincents Institute VIC Australia
Paul Wang ACRF South Australian Cancer SA Australia
Genomics Facility, Centre For Cancer
Biology, SA Pa
Trent Warburton TrendBio Pty Ltd VIC Australia
Paul Waring The University Of Melbourne VIC Australia
Emma Whitelaw La Trobe University VIC Australia
Liam Williams University Of Auckland New Zealand
Trevor Wilson Monash Health Translation Precinct VIC Australia
Barbara Wold California Institute of Technology USA
Stephen Wong Peter MacCallum Cancer Centre VIC Australia
Janice Woods Millennium Science VIC Australia
James Wynne CSIRO Biosecurity Flagship VIC Australia
Monika Zavodna Otago Genomics & Bioinformatics New Zealand
Facility, Uni Of Otago
Liang Zhang Chinese National Human Genome China
Center At Shanghai
Kerong Zhang The Australian National University ACT Australia
Li Zhou The Tumour Bank At The Children’s NSW Australia
Hospital At Westmead
137
Handbook
The 2015 AGTA conference
will cover themes that include cancer genomics,
epigenomics, bioinformatics, transcriptomics, functional
genomics, clinical sequencing and plant genomics with
Save
a special focus on the genomics of good wine. We have
invited an outstanding list of international and national
speakers plus have engaged industry to showcase the best
in new genomic technologies. Please come and join us for
the Date!
some great networking and cutting edge science in the
heart of Australia’s top food and wine country.
We look forward to seeing you at AGTA 2015 in the Hunter
Valley to share the excitement of cutting-edge science!
11-14 October 2015
Co-Convenors
Marcel Dinger
Garvan Institute of Medical Research, Sydney
carsten Kulheim
Australian National University, Canberra
Any Questions?
Contact Brigitte at Leishman Associates
brigitte@leishman-associates.com.au
Phone: +61 3 6234 7844
AGTA 2014
SPONSORS AND EXHIBITORS
On Q
software

2014 AGTA Conference Handbook

Uploaded by

Copyright:

Available Formats

2014 AGTA Conference Handbook

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

2014 AGTA Conference Handbook

Uploaded by

Copyright:

Available Formats

What is the purpose of the conference?

What is the purpose of the conference?

What topics will be covered at the conference?

What topics will be covered at the conference?

HANDBOOK

Dr Ruby C Y Lin (President), Asbestos Diseases

Dr Jac Charlesworth, Menzies Research Dr Mark Waltham, St Vincent’s Institute of

Associate Professor Daniel Catchpoole, Dr Noel Cogan, Department of Environment

ARC to consult with us for all things ‘omics”.

WELCOME FROM THE

VISIT THE MILLENNIUM SCIENCE BOOTH TO LEARN MORE

Registration Desk Conference Proceedings

Wednesday 15 October 2014 Username: AGTA1

Photographs, Videos, Recording

© 2014 Illumina, Inc. All rights reserved. Images

1400 – 1800 Registration Open Promenade

OPENING ORATION Promenade 1

Chairs: Dr Ruby CY Lin

KEYNOTE SPEAKER INVITED SPEAKER STUDENT EARLY CAREER RESEARCHER SPONSORED

0730 – 1715 Registration Desk Open Promenade Registration

0800 – 1715 Trade Exhibition Open Promenade Foyer &

0830 – 0845 Official Welcome and Conference Opening Promenade 1

SESSION 1: PLANT GENOMICS Promenade 1

Chairs: Dr Carsten Kulheim and Dr Noel Cogan

SESSION 2: EPIGENETICS Promenade 1

Chairs: Professor Emma Whitelaw and Dr David Martino

KEYNOTE SPEAKER INVITED SPEAKER STUDENT EARLY CAREER RESEARCHER SPONSORED

1245 – 1300 Widespread epigenomic and gender-specific

1330 – 1515 Poster Session 1 Promenade Foyer &

SESSION 3: BIOINFORMATICS Promenade 1

Chairs: Professor Gordon Smyth and Dr David Goode

KEYNOTE SPEAKER INVITED SPEAKER STUDENT EARLY CAREER RESEARCHER SPONSORED

0800 – 1730 Registration Desk Open & Welcome Refreshments Promenade

0800 – 1730 Trade Exhibition Open Promenade Foyer &

0825 – 0830 Welcome to Day Two Promenade 1

Proudly sponsored by:

SESSION 5: CLINICAL SEQUENCING Promenade 1

Chairs: Ms Vivien Vasic and Professor Paul Waring

KEYNOTE SPEAKER INVITED SPEAKER STUDENT EARLY CAREER RESEARCHER SPONSORED

1300 – 1500 Lunch, Poster session 2 & Trade Exhibition

Meeting Room 3 & 4

CLINICAL SEQUENCING WORKSHOP

1550 – 1615 SAMPLE PREPARATION AND SEQUENCING

SESSION 6: TRANSCRIPTOMICS (1500 – 1730)

KEYNOTE SPEAKER INVITED SPEAKER STUDENT EARLY CAREER RESEARCHER SPONSORED

1615 – 1645 Unraveling Dendritic Cell-Subset Proudly sponsored by:

Commitment - One Cell At A Time

KEYNOTE SPEAKER INVITED SPEAKER STUDENT EARLY CAREER RESEARCHER SPONSORED

0830 – 1530 Registration Desk Open & Welcome Refreshments Promenade

0830 – 1530 Trade Exhibition Open Promenade Foyer &

0855 – 0900 Welcome to Day Three Promenade 1

SESSION 7: FUNCTIONAL GENOMICS

SESSION 8: SINGLE MOLECULE SEQUENCING

Insights into Human Genome

1300 – 1345 Lunch & Trade Exhibition Promenade Foyer &

KEYNOTE SPEAKER INVITED SPEAKER STUDENT EARLY CAREER RESEARCHER SPONSORED

SESSION 9: METAGENOMICS Promenade 1

Chairs: Dr Torsten Seemann and Dr Mark Schultz

2015 Conference Launch

KEYNOTE SPEAKER INVITED SPEAKER STUDENT EARLY CAREER RESEARCHER SPONSORED