No. Name Affiliation Department
1 Nasser Mohieddin Abukhdeir University of Waterloo Chemical Engineering
Title: Pattern Recognition and Characterization for Imaging of Self-Assembly Processes
Abstract: R. J. Suderman^1 , N. M. Abukhdeir^2 , and D. Lizotte^1 ^1 Faculty of Mathematics, David R. Cheriton School of Computer Science, University of Waterloo, Waterloo, Ontario ^2 Department of Chemical Engineering, University of Waterloo, Waterloo, Ontario Modern microscopy and spectroscopy techniques have enabled increasingly detailed imaging of self-assembled materials and the dynamics of their formation. Self-assembled materials exhibit a phase-ordering transition, typically from a disordered phase to an ordered phase with some degree of orientational or translational order. These phenomena have motivated research into "bottom-up" manufacturing methods which leverage self-assembly to form surfaces and volumes with desired nanoscale structure. A key issue in the development of such methods is that there currently exists few methods to rigorously characterize the structure and dynamics of these self-assembled domains. Up until now, analysis of self-assembly imaging data has been determined either in a purely qualitative way, or at best using simple heuristic algorithms. In this work an image analysis methodology is presented which enables self-assembly image (1) pattern detection and (2) classification. The presented methodology uses the discrete Fourier transform to both determine if patterns are present in the imaged surface and to guide identification of self-assembled regions within the image. Localized image analysis is then performed using a wavelet transform which identifies individual patterned regions within the image. Finally, for a sample surface containing both hexagonal and striped regions, patterns are classified using bond-orientational order theory. The results of this work provide a significant first step towards enabling quantitative analysis of self-assembly imaging data.
2 Dalibor Dvorski Wilfrid Laurier University Physics & Computer Science
Title: Julia sets computation using high performance computing
Abstract: In the early twentieth century, Gaston Maurice Julia studied the iteration of polynomials and rational functions. Interesting results may be observed by iterating some x in f(x). If function iteration is extended to complex function iteration then we may define a Julia set as being generated by Qc(z)=z^2+c, where c is a complex number of the form a+bi, and where a and b are real numbers. Some resulting images of Julia sets show great beauty of contemporary art. For such beauty, high performance computing is utilized to determine the points of a Julia set. We present a project, written using C, MPI, and OpenGL, that uses the available parallel-processing power of SHARCNET to compute Julia sets and display its resulting images.
3 Abbas Golestani University of Windsor School of Computer Science
Title: Speciation with gene flow in a heterogeneous virtual world: can physical obstacles accelerate speciation?
Abstract: The origin of species remains one of the most controversial and least understood topics in evolution. While it is widely accepted that complete cessation of gene-flow between populations due to long-lasting geographic barriers results in a steady, irreversible increase of divergence and eventually speciation, the extent to which various degrees of habitat heterogeneity influences speciation rates is less well understood. Here, we investigate how small, randomly distributed physical obstacles influence the distribution of populations and species, the level of population connectivity (e.g., gene flow) as well as the mode and tempo of speciation in a virtual ecosystem composed of prey and predator species. We adapted an existing individual-based platform, EcoSim, to allow fine tuning of the gene flow's level between populations by adding various numbers of obstacles in the world. The platform implements a simple food chain consisting of primary producers, herbivores (prey) and predators. It allows complex intra- and inter-specific interactions, based on individual evolving behavioral models, as well as complex predator-prey dynamics and coevolution in spatially homogenous and heterogeneous worlds. We observed a direct and continuous increase in the speed of evolution (e.g., the rate of speciation) with the increasing number of obstacles in the world. The spatial distribution of species was also more compact in the worlds with obstacles than in the world without obstacles. Our results suggest that environmental heterogeneity and other factors affecting demographic stochasticity can directly influence speciation and extinction rates.
4 Morteza Mashayekhi University of Windsor Computer Science
Title: Investigating the Effect of Spatial Distribution and Spatiotemporal Information on Speciation using Individual-Based Ecosystem Simulation
Abstract: In this paper, we investigate the impact of species’ spatial and spatiotemporal distribution information on speciation, using an individual-based ecosystem simulation (Ecosim). For this purpose, using machine learning techniques, we try to predict if one species will split in near future. Because of the imbalanced nature of our dataset we use smote algorithm to make a relatively balanced dataset to avoid dismissing the minor class samples. Experimental results show very good predictions for the test set generated from the same run as the learning set. It also shows good results on test sets generated from different runs of Ecosim. We also observe superior results when we use, for the learning set, a run with more species compare to a run with less species. Finally we can conclude that spatial and spatiotemporal information are very effective in predicting speciation.
5 Konstantinos Parsopoulos Wilfrid Laurier University Physics & Computer Science
Title: Ant-Based Approaches for Solving Autocorrelation Problems
Abstract: The present paper proposes two different ant-based formulations for tackling autocorrelation problems that frequently appear in Computer Science applications. The formulations are combined with different ACO variants, producing a variety of algorithms. Experimental tests are performed on two problems of different dimensionality and degree of difficulty. The proposed approaches are statistically compared against a specialized TS approach with verified efficiency. Useful conclusions are derived and questions that require further investigation arise.
6 Alex Razoumov University of Ontario Institute of Technology Science
Title: SHARCNET visualization help
Abstract: SHARCNET staff can help researchers with visualization of their large datasets. In addition to an array of standalone workstations, we have a number of visualization packages installed on our systems, and we can provide expert help from importing data to interactive rendering. Here we describe several visualization projects we worked on recently in collaboration with SHARCNET researchers.
7 Silvija Smith Wilfrid Laurier University Chemistry
Title: A density functional theory study of hydrogen transfer for short-chain alkane thiols on small cationic and neutral gold clusters
Abstract: To prevent aggregation of gold clusters in solution, sulphur containing molecules are often used. The bonds formed between sulphur and gold are extremely strong and, consequently, monolayers of thiols form at the surface of gold clusters, protecting it from attack by other functional groups present in solution. Thiols consisting of long-chain alkanes form a densely packed monolayer on the surface of the gold clusterfurther stabilizing the complex and preventing aggregation. In this study, density functional theory was used with the BP86 functional to determine the optimized geometries and vibrational frequencies of cationic and neutral gold cluster-thiol complexes. These complexes consist of 1-4 gold atoms bonded to hydrogen sulfide, H2S, methane thiol, CH3SH, and ethane thiol, CH3CH2SH. A pseudopotential was employed to account for the large relativistic effects of gold. The thermochemical values of these gold cluster-thiolate complexes were also calculated, to determine whether the transfer of hydrogen was a spontaneous process. Significantly different results were obtained for the cationic and neutral species of these short-chain alkane thiol-gold cluster complexes.
8 Brad Van Oosten Brock University Physics
Title: Small molecule interactions with lipid bilayers: a molecular dynamics study of chlorhexidine
Abstract: Chlorhexidine is a chemical antiseptic shown to be effective against a wide range of bacteria. It presents an interesting modelling challenge with a hydrocarbon chain connecting two biguanides (arginine analogues) and two aromatic rings. We constructed all-atom force field parameters for chlorhexidine derived from the CHARMM36 force field using well established bonded parameters of certain amino acids. Partial charges were treated differently, which were obtained quantum mechanically with DTF calculations using the GAUSSIAN software on SHARCNET. The SHARCNET systems were also used to conduct molecular dynamic simulations using the GROMACS simulation software to reproduce the experimental environment of chlorhexidine in a 1,2-Dimyristoyl-sn-Glycero-3-Phosphocholine (DMPC) bilayer to produce atomic-level information. We will compare and contrast the results of our model to that of the neutron scattering experiments previously done in our lab.
9 Marek Wartak Wilfrid Laurier University Physics
Title: Simulation of semiconductor laser using nonequilibrium Green’s functions
Abstract: Simulation of semiconductor laser using nonequilibrium Green’s functions J. Miloszewski and M.S. Wartak We present novel method to simulations of quantum well, edge-emitting semiconductor lasers based on non-equilibrium Green's functions (NEGF) method. We illustrated our approach by considering steady-state analysis with fully quantum mechanical description of carrier and photon dynamics. General equations for photon Green's functions and polarizations are derived as well as their coupling to electron Green's functions through self-energies. Series of approximations are done to photon Green's function to allow for efficient numerical approach. All equations are written in the non-orthogonal basis suitable for numerical calculations. As an example, the theory is applied to analyze AlGaAs\AlAS quantum well laser with the effective mass Hamiltonian. Major laser characteristics such as material and modal gain, threshold parameters, carrier and current densities were determined.
Title: Calculating Quantum Modes using a Finite Difference Technique
Abstract: Calculating Quantum Modes using a Finite Difference Technique Jacek Miloszewski and Andrew McNeely A quantum dot is a portion of matter confined in three dimensions. Crystal growth technologies, such as metal-organic chemical vapor deposition (MOCVD) and molecular beam epitaxy (MBE), enable the growth of quantum dots with various geometries for a diverse field of applications. Due to the ease of tailoring and the many applications in fields such as optoelectronics and quantum computing, there is great interest in accurate solution methods of band structures within the quantum dot for device design and performance evaluation. We used the finite difference method which provides a computationally efficient and flexible tool for investigating the behaviour of quantum dot systems. Some preliminary results are presented.
10 Tyson Whitehead University of Western Ontario SHARCNET
Title: Colloidal Particles Interacting in the Presence of a Liquid Crystal
Abstract: NOTE: The authors of this poster are Frances Mackay and Colin Denniston. I am presenting it as I have been working with the PI to port their lattice-Boltzmann code to the GPUs as part of the dedicated programming (work in progress). ABSTRACT: When immersed in a liquid crystal, colloidal particles generate defects and distortions which lead to significantly different particle interactions compared to those arising in an isotropic fluid. Here, we use a lattice-Boltzmann algorithm to numerically investigate these interactions, for a variety of situations. We present results for a defect-bounded chain formed in a cholesteric liquid crystal, and investigate the behavior of 2D deformable particles interacting in a nematic liquid crystal.
Title: Revisiting BIGMIN on Z-order curves
Abstract: The Z-order curve maps multidimensional data (e.g., spacial data) onto a single dimension (e.g., memory) in a way that helps preserve locality. This sort of mapping is critical for achieving optimal performance in locality sensitive algorithms. The BIGMIN calculation computes the next point of intersection of the Z-order curve with a hyper-box. It is a critical component for dealing with sparse Z-order curves. The original BIGMIN computation is a recursive algorithm which, while sufficient fast for working with disk based data, represents a significant portion of computation time when working with RAM based data. We present a new seven-line constant-time BIGMIN computation based on parallel bit operations to resolve this issue.

© 2013 Shared Hierarchical Academic Research Computing Network (www.sharcnet.ca).