The ATLAS experiment collects data on the world's highest energy proton-proton collisions. Its primary goals are to search for direct evidence of physics beyond the Standard Model of particle physics, and to provide a final verdict on the existence of the Standard Model Higgs boson. A robust τ identification procedure is an important ingredient in the search for Standard Model and supersymmetric Higgs bosons. In this study, new τ identification variables based on calorimeter topological clusters are investigated. These variables complement existing calorimeter-based variables by bringing additional discriminating power against hadronic jets. They have a straightforward physics interpretation which allows them to be calculated at particle-level. Some of these new variables are shown to be robust against background noise from the underlying event and in-time pileup, two effects which negatively impact τ identification. Several of these new variables are now part of the official τ reconstruction software at ATLAS. iii À ceux et celles qui m'encouragent et qui n'ont jamais douté de moi malgré mes propres doutes: Papa, Maman, Mirka, Maxime, mes amis de toujours et Elisabeth. iv

Abstract Pulsed fields are widely used in gel electrophoretic separations to increase resolution. For instance, Boyde & To [1] presented experimental results for the separation of spherical particles using pulsed fields. They first used alternating fields of fixed amplitudes () applied in the forward direction for a duration twice as long as in the backward direction. They then used field interruption in order to allow particles to thermally detrap. In both cases, they reported that using a pulsed field makes the particles migrate faster.

This report is a .study of the dynamic properties of semiconductor laser diodes. The measurement of some important lc;ser diode parameters necessary for dyn?mic behaviour prediction is described. The relc;xation oscillation behaviour for laser diodes pumped with nanosecond time scale current pulses is predicted using both an approximate an~lytic solution c;nd computer simulations. This predicted behaviour is compared with experimental results. DynC~mic experiments with c;n externc;l cavity for extra optical feedb2ck and a regenerativE loop for optoelectronic feedback are also described and discussed. Details of the experimental setups and techniques used are given. iii and B.L.C. for their support. Special thanks to John Goodwin for numerous discussions and immeasurable help throughout this work.

retical curve and experimental results (points) obtained for the system sensitivity as a function of the angle between the acoustic beam propagation direction and the line joining the foci. There is a general agreement between the two. Fig. 3a shows the acoustic beam profile (amplitude) within the Fresnel region of the transducer, in good agreement with its theoretical counterpart (Fig. 3b).

This report is a .study of the dynamic properties of semiconductor laser diodes. The measurement of some important lc;ser diode parameters necessary for dyn?mic behaviour prediction is described. The relc;xation oscillation behaviour for laser diodes pumped with nanosecond time scale current pulses is predicted using both an approximate an~lytic solution c;nd computer simulations. This predicted behaviour is compared with experimental results. DynC~mic experiments with c;n externc;l cavity for extra optical feedb2ck and a regenerativE loop for optoelectronic feedback are also described and discussed. Details of the experimental setups and techniques used are given. iii and B.L.C. for their support. Special thanks to John Goodwin for numerous discussions and immeasurable help throughout this work.

Using Arnold's Classification Theorem applied to a four-dimensional manifold, it is shown that there is only a finite number of ways in which energy can discontinuously change state. It is demonstrated that each of these energy flow pathways can be associated with a distinct elementary particle. The theory not only shows how the formation of particles from the stress-energy present in the space-time manifold can be predicted from first principles, but also that there must exist five fundamental forces in a universe in which discontinuous energy transitions are possible. Finally, the existence of a new, as yet undiscovered particle is predicted, which is associated with this new fifth force.

Molecular spiders are synthetic molecular motors featuring multiple legs that each can interact with a substrate through binding and cleavage. Experimental studies suggest the motion of the spider in a matrix is biased toward uncleaved substrates and that spider properties such as processivity can be altered by changing the binding strength of the legs to substrate ͓R. Pei, S. K. Taylor, D. Stefanovic, S. Rudchenko, T. E. Mitchell, and M. N. Stojanovic, J. Am. Chem. Soc. 128, 12693 ͑2006͔͒. We investigate the origen of biased motion and molecular motor properties of bipedal spiders using Monte Carlo simulations. Our simulations combine a realistic chemical kinetic model, hand-over-hand or inchworm modes of stepping, and the use of a onedimensional track. We find that stronger binding to substrate, cleavage and spider detachment from the track are contributing mechanisms to population bias. We investigate the contributions of stepping mechanism to speed, randomness parameter, processivity, coupling, and efficiency, and comment on how these molecular motor properties can be altered by changing experimentally tunable kinetic parameters.

that measure chemotaxis as a ''yes/no'' response, relative degrees of response to molecular stimuli and negative responses can also be measured within this microfluidic device. Movement of the predatory bacteria within these gradients should provide critical information about the molecules responsible for Bdellovibrio chemotaxis toward prey.

The rational design of peptides that fold to form discrete nanoscale objects, and/ or self-assemble into nanostructured materials is an exciting challenge. Such efforts test and extend our understanding of sequence-to-structure relationships in proteins, and potentially provide materials for applications in bionanotechnology. Over the past decade or so, rules for the folding and assembly of one particular protein-structure motif-the a-helical coiled coilhave advanced sufficiently to allow the confident design of novel peptides that fold to prescribed structures. Coiled coils are based on interacting a-helices, and guide and cement many protein-protein interactions in nature. As such, they present excellent starting points for building complex objects and materials that span the nano-to-micron scales from the bottom up. Along with others, we have translated and extended our understanding of coiled-coil folding and assembly to develop novel peptide-based biomaterials. Herein, we outline briefly the rules for the folding and assembly of coiled-coil motifs, and describe how we have used them in de novo design of discrete nanoscale objects and soft synthetic biomaterials. Moreover, we describe how the approach can be extended to other small, independently folded protein motifs-such as zinc fingers and EF-handsthat could be incorporated into more complex, multi-component synthetic systems and new hybrid and responsive biomaterials.

Molecular spiders are synthetic molecular motors featuring multiple legs that each can interact with a substrate through binding and cleavage. Experimental studies suggest the motion of the spider in a matrix is biased toward uncleaved substrates and that spider properties such as processivity can be altered by changing the binding strength of the legs to substrate ͓R. Pei, S. K. Taylor, D. Stefanovic, S. Rudchenko, T. E. Mitchell, and M. N. Stojanovic, J. Am. Chem. Soc. 128, 12693 ͑2006͔͒. We investigate the origen of biased motion and molecular motor properties of bipedal spiders using Monte Carlo simulations. Our simulations combine a realistic chemical kinetic model, hand-over-hand or inchworm modes of stepping, and the use of a onedimensional track. We find that stronger binding to substrate, cleavage and spider detachment from the track are contributing mechanisms to population bias. We investigate the contributions of stepping mechanism to speed, randomness parameter, processivity, coupling, and efficiency, and comment on how these molecular motor properties can be altered by changing experimentally tunable kinetic parameters.

Molecular spiders are synthetic biomolecular walkers that use the asymmetry resulting from cleavage of their tracks to bias the direction of their stepping motion. Using Monte Carlo simulations that implement the Gillespie algorithm, we investigate the dependence of the biased motion of molecular spiders, along with binding time and processivity, on tunable experimental parameters, such as number of legs, span between the legs, and unbinding rate of a leg from a substrate site. We find that an increase in the number of legs increases the spiders’ processivity and binding time but not their mean velocity. However, we can increase the mean velocity of spiders with simultaneous tuning of the span and the unbinding rate of a spider leg from a substrate site. To study the efficiency of molecular spiders, we introduce a time-dependent expression for the thermodynamic efficiency of a molecular motor, allowing us to account for the behavior of spider populations as a function of time. Based on this definition, we find that spiders exhibit transient motor function over time scales of many hours and have a maximum efficiency on the order of 1%, weak compared to other types of molecular motors.

Molecular spiders are synthetic molecular motors featuring multiple legs that each can interact with a substrate through binding and cleavage. Experimental studies suggest the motion of the spider in a matrix is biased toward uncleaved substrates and that spider properties such as processivity can be altered by changing the binding strength of the legs to substrate ͓R. Pei, S. K. Taylor, D. Stefanovic, S. Rudchenko, T. E. Mitchell, and M. N. Stojanovic, J. Am. Chem. Soc. 128, 12693 ͑2006͔͒. We investigate the origen of biased motion and molecular motor properties of bipedal spiders using Monte Carlo simulations. Our simulations combine a realistic chemical kinetic model, hand-over-hand or inchworm modes of stepping, and the use of a onedimensional track. We find that stronger binding to substrate, cleavage and spider detachment from the track are contributing mechanisms to population bias. We investigate the contributions of stepping mechanism to speed, randomness parameter, processivity, coupling, and efficiency, and comment on how these molecular motor properties can be altered by changing experimentally tunable kinetic parameters.

Molecular spiders are synthetic molecular motors featuring multiple legs that each can interact with a substrate through binding and cleavage. Experimental studies suggest the motion of the spider in a matrix is biased toward uncleaved substrates and that spider properties such as processivity can be altered by changing the binding strength of the legs to substrate [R. Pei, S. K. Taylor, D. Stefanovic, S. Rudchenko, T. E. Mitchell, and M. N. Stojanovic, J. Am. Chem. Soc. 128, 12693 (2006)]. We investigate the origen of biased motion and molecular motor properties of bipedal spiders using Monte Carlo simulations. Our simulations combine a realistic chemical kinetic model, hand-over-hand or inchworm modes of stepping, and the use of a one-dimensional track. We find that stronger binding to substrate, cleavage and spider detachment from the track are contributing mechanisms to population bias. We investigate the contributions of stepping mechanism to speed, randomness parameter, processivity, coupling, and efficiency, and comment on how these molecular motor properties can be altered by changing experimentally tunable kinetic parameters.

The mechanical properties of the cell’s structural components influence the size, shape, and functionality of the cell throughout its division cycle. For example, a combination of the plasma membrane’s edge tension and bending resistance sets a lower bound on cell size, while the membrane’s tear resistance sets a pressure-dependent upper bound on the size of cells lacking a cell wall. The division cycle of the simplest cells may be dominated by one or two principles such as the maximization of entropy, or the minimization of energy or structural materials. By studying colonies of cells, modern and fossilized, with techniques from classical and statistical mechanics, a partial history can be charted for the appearance and properties of the simplest cell designs.

A novel method to screen for transcription factors binding to promoter DNA sequences has been developed using DNA chip surfaces and mass spectrometry. This technique was demonstrated with Escherichia coli lac repressor, LacI. The consensus promoter binding sequence for LacI and a scrambled version of the same DNA sequence were prepared on two affinity chip surfaces. Total E. coli protein lysate was applied to the two surfaces. A 38.2 kDa protein, as detected by SELDI-MS, was captured on the chip surface containing the binding sequence for LacI but not on the surface containing the scrambled sequence. The protein was identified following one-step, small-scale affinity capture and peptide mapping. Subsequent database searches identified the 38.2 kDa protein as the lac repressor of E. coli. We discuss application of DNA chip affinity capture to characterize transcription factors and to screen for differences in cellular regulatory networks. © 2002 Elsevier Science (USA)

Synaptotagmin I is a critical component of the synaptic machinery that senses calcium influx and triggers synaptic vesicle fusion and neurotransmitter release. Fluorescence resonance energy transfer studies conducted on synaptotagmin demonstrate that calcium concentrations required for fusion induce a conformational change (EC50 3 mM) that brings the two calcium-binding C2 domains in synaptotagmin closer together. Analytical ultracentrifugation studies reveal that synaptotagmin is monomeric under these conditions, indicating that this calcium-triggered association between theC2 domains is intramolecular, rather than intermolecular. These results suggest a mechanism for synaptotagmin function at the presynaptic plasma membrane that involves the self-association of C2 domains.