Combining Prior Knowledge, Graph Theory and Multivariate Experiments to Reverse Engineer Signaling Networks in Mammalian Cells

April 4, 2007
Time: 11:00am-12:00pm
EE Conference Room 1300 Mudd
Speaker: Avi Ma'ayan, Mount Sinai School of Medicine

Abstract

Developments in biochemistry and molecular biology in the past thirty years produced an impressive list of cellular components and interactions in mammalian cells. Since this data is stockpiled in natural text, organization of such knowledge into networks is necessary. Topology analysis of such networks uncovered interesting global properties. At the same time, advancements in experimental techniques to measure and visualize many cellular components at once grow in diversity and accuracy. Multivariate data produced by experiments can be combined with prior knowledge from legacy literature to develop graph theory inspired models of cellular systems made of interacting cellular components. Experimental results are analyzed using networks created from literature to make predictions about the role of undiscovered components and pathways. This approach has been successful for quality assessment of pre-synaptic proteomics data; it enabled us to discover new pathways and components in cannabinoid induced neurite outgrowth using transcription-factor arrays in combination with pharmacological inhibitors and RNAi; and assisted in identifying a novel Noonan Syndrome causing gene. Combining literature-based networks and graph theory with multivariate experiments can rapidly enhance our understanding of the detail workings of mammalian cells.


500 W. 120th St., Mudd 1310, New York, NY 10027    212-854-3105               
©2014 Columbia University