Scheduled Speakers

Donald Stuss

Matthew Norton

Shiva Amiri

Igor Jurisica

Brendan Johns

Khaled El Emam

Peter Snelling

Igor Jurisica, Senior Scientist Ontario Cancer Institute

Igor JurisicaIgor Jurisica is a Canada Research Chair in Integrative Computational Biology, focusing on cancer research. To significantly impact cancer research, novel therapeutic approaches for targeting metastatic disease and diagnostic markers reflective of changes associated with disease onset that can detect early stage disease must be discovered. Better drugs must be rationally designed, and current drugs made more efficacious either by re-engineering or by information-based combination therapy. To tackle these complex biological problems and impact high-throughput biology requires integrative computational biology, i.e., considering multiple data types, developing and applying diverse algorithms for heterogeneous data analysis and visualization. Improved analysis and reasoning algorithms will in turn advance disease diagnosis by finding better markers, and improve patient management by supporting information-based medicine. Combined, this will 1) advance computational algorithms; 2) help to fathom cancer biology; and 3) lead to creating computational models of cancer.

Igor Jurisica is a Senior Scientist at the Ontario Cancer Institute, University Health Network, Associate Professor in the Departments of Computer Science and Medical Biophysics, University of Toronto, Adjunct Professor at the School of Computing,, Queen's University, and a Visiting Scientist at theIBM Centre for Advanced Studies. Dr. Jurisica's research focuses on integrative computational biology, and representation, analysis and visualization of high dimensional data generated by high-throughput biology experiments, in the context of Cancer Informatics. Of particular interest is the use of comparative analysis for the mining of integrated different datasets such as protein-protein interaction, gene and protein expression profiling, and high-throughput screens for protein crystallization.

Title:

High-performance computing and "big data" in integrative cancer informatics. From biomarkers to new drugs and increased survival

Abstract:

Precision medicine requires better identification of patient subgroups and new treatment options. Single drugs are rarely sufficient; thus, we need to find drug combinations. While drug mechanisms-of-action are complex and poorly understood, integrating computational approaches with advanced protein interaction mapping and drug targeting will enable us to predict, characterize and validate novel therapeutic approaches.

Data on thousands of cancer patient profiles from diverse technology platforms provide essential resources for molecular medicine. However, effectively integrating, annotating, and analyzing these high-dimensional, heterogeneous and distributed data with aim to create intelligent hypotheses and realistic models of human disease is not trivial.

Integrative computational and network-based analysis can help to unravel mechanism of action for therapeutics, re-position existing drugs for novel use and prioritize multiple candidates based on predicted toxicity, identify groups of patients that may benefit from treatment and those where a given drug would be ineffective. We need to integrate algorithms and systems from machine learning, databases, image and text analysis, ontologies, human-computer interaction, graph theory and visualization to tackle these diverse problems.

Scale changes everything, and size does matter. Integrating intelligent heuristics with novel computing environments, such as grid and GPU computing, provides scalable platform for these big data challenges.

Protein-protein interaction network-based analysis of diverse drug-related datasets - including drug targets, compound structures, drug-regulated genes - confirms strong relationships between network structure and protein function, and compound structure and mechanism of action. As expected, drug-regulated genes differ from drug targets in terms of functional annotations, cellular localizations, and topological properties. Drug targets mainly include receptors on the plasma membrane, down-regulated genes are largely in the nucleus and are enriched for DNA binding, and genes lacking drug relationships are enriched in the extracellular region. Network topology analysis indicates several significant graph properties, including high degree and betweenness for the drug targets and drug-regulated genes. Topological analysis suggests that proteins of down-regulated genes are frequently involved in complexes. Analyzing network distances between regulated genes shows that genes regulated by structurally similar drugs are significantly closer than genes regulated by dissimilar drugs. Finally, network centrality of a drug.s differentially regulated genes correlates significantly with drug toxicity.