Moving from Correlation to Causation using Biological Pathways

Wednesday, February 29, 2012 - 12:00pm - 12:15pm
Keller 3-180
Gary Bader (University of Toronto)
Human genetics research traditionally identifies genes underlying a phenotype using association (correlation) analysis. A gene mutation within a population is associated with a phenotype, such as a disease. Biological pathways represent knowledge about molecules, processes and their interactions. Maps of such pathways are used to design and analyze experiments, and for predicting the behavior of biological systems. Pathway information can help explain how a genetic perturbation leads to an altered phenotype, and thus helps move from association (correlation) at the genetic variant or gene levels to mechanism (causality) at the cellular and physiological levels. Pathways are represented as networks, but capture process information (one step leads to another). The increasing amount of pathway information available in public databases represents a major opportunity for network research, as network analysis algorithms must be adapted to pathway information. We are building a number of database and software resources to help access, visualize and analyze pathway information, including Pathway Commons as a convenient single point of access to diverse biological pathway information translated to a common data language (BioPAX), GeneMANIA and Cytoscape. These collaborative projects are important steps towards the development of a complete and integrated computable map of the cell across all species and developmental stages.