Download code from https://github.com/xa6xa6/metaOthello/

There are two code folders:

• Builder -- The code under "build" is used to generate the MetaOthello index.
• Classifier -- The code under "classifier" is used to perform taxonomic classification of sequencing reads using MetaOthello Index.

Building indexing is very time-consuming (costs about 6 hours to build index for NCBI/refseq bacterial genome database). Therefore, we provide ready-built indexes (NCBI/refseq bacterial genome database) for users to download:

• 20mer index: https://drive.google.com/open?id=0BxgO-FKbbXRIYWREa2NwejlVYUU
• 25mer index: https://drive.google.com/open?id=0BxgO-FKbbXRIY1pRaHJsYVg5dTQ
• 31mer index: https://drive.google.com/open?id=0BxgO-FKbbXRIa0Flc3Q4bWtycGM

If you want to build an index with your own reference sequences, follow these steps.

Jellyfish is used to prepare k-mer files for each reference sequence. Download jellyfish from: http://www.cbcb.umd.edu/software/jellyfish/

1. Produce a k-mer count file for your reference seqeuences. Command:

    jellyfish count \
    –o <path_to_bacterial_rawKmerCountFile> \
    -m <Kmer_length> \
    -t <threads_num> \
    -s <bf_size> \
    -C <path_to_bacterial_referenceSeqFastaFile>

2. Dump k-mers to human-readable format. Command:

    jellyfish dump \
    –t –c \
    –o <path_to_bacterial_readableKmerCountFile> \
    <path_to_bacterial_rawKmerCountFile>

3. Generate taxonomy info file. Put all readable k-mer count files into the same directory <path_to_bacterial_reference_seq_Kmer_file_dir> and rename them as 1.Kmer, 2.Kmer, ..., m.Kmer, and generate a taxonomy info file like: https://drive.google.com/open?id=0BxgO-FKbbXRIZlV3ZzBBdlFpMTQ

   There are three columns for each taxonomic rank in the file: the 1st column is a reissued id from 0 to m-1, where m is the total taxon num in that taxonomic rank. The 2nd column lists taxon ids, and the 3rd column lists taxon scientific names. Each row represents a species and its associated taxonomy info.

4. Run make build under the directory build.

5. Build the index. Command:

    ./build \
    <bacterial_reference_seq_associated_taxonomy_info_file(generated in Step1.3)> \
    <path_to_bacterial_reference_seq_Kmer_file_dir> \
    <shared kmer file suffixes> \
    <Kmer_length> 6 \
    <path_to_bacterial_index> \
    <path_to_a_temp_dir_for_intermediate_files>

1. Run make classifier in the classifier directory.
2. Perform taxonomic classification for each metagenomics sequencing reads. Command:

    ./classifier \
    <path_to_bacterial_index> \
    <path_to_output_results_dir> \
    <Kmer_length> \
    <threads_num> \
    <fa_or_fq> \
    <SE_or_PE> \
    <bacterial_speciesId2taxoInfo_file> \ 
    <NCBI_names_file> \
    <readFile_singleEnd or readFile_end1> \
    (<readFile_end2 if paired-end reads are provided>)

<bacterial_speciesId2taxoInfo_file> can be downloaded from: https://drive.google.com/open?id=0BxgO-FKbbXRIc3FkLVFvMlpVVGM

Each row represents a species and its associated taxon ids at each taxonomic rank: species, genus, family, order, class, and phylum. Assign -1 if the taxon id is not available.

<NCBI_names_file> can be downloaded from: https://drive.google.com/open?id=0BxgO-FKbbXRIUFI2dHlBMXZhdTA

NOTE: We will keep the following files updated with the latest NCBI/refseq bacterial genome databases:

1. bacterial reference seq associated taxonomy info file,
2. bacerial index (MetaOthello index for classification)
3. bacterial speciesId2taxoInfo_file
4. NCBI names file

Also, we will release tools for generating all the above files (from NCBI/refseq bacterial genome databases) very soon.

Copyright (C) 2016-, University of Kentucky
Please refer to LICENSE.TXT for the detailed 'License'.