Predicting Cancer-specific Vulnerability via Data-driven Detection of Synthetic Lethality
Wednesday, October 15, 2014
10:00 a.m.-11:00 a.m.
FDA-White Oak Bldg. 2 Room 2047E
Eytan Ruppin, MD, Ph.D.
Director of the Center for Bioinformatics and Computational Biology, UMD
Eytan Ruppin received his MD (1987) and Ph.D. in Computer Science (1993) from Tel-Aviv University, Israel, where he has been a Professor of Medicine and Computer Science. In July 2014 he has joined the Dept. of Computer Science at the University of Maryland as a professor and director of its center for bioinformatics and computational biology (CBCB). Starting his career his in computational neuroscience his research has later shifted to computational systems biology with emphasis on genome scale metabolic modeling (GSSM) of human, microbial and plant metabolism. Studying human metabolism, his lab published the first large scale computational description of human metabolism across different tissues (Nature Biotechnology 2008), the first model of the liver metabolism (MSB 2010) and have utilized GSSM to predict biomarkers of metabolic disorders (MSB 2009). It developed the first generic model of cancer metabolism (MSB 2011), which has led to the successful identification of a new synthetic lethality based cancer drug target (Nature 2011) and the recent publication of personalized cell-specific cancer metabolic models (MSB 2014).
Synthetic lethality occurs when the inhibition of two genes is lethal while the inhibition of each individual gene is not. It can be harnessed to selectively treat cancer by identifying genes that are inactive in a given cancer and targeting their Synthetic Lethal (SL)-partners. We present a data-driven computational pipeline for the genome-wide identification of candidate SL-interactions in cancer, by analyzing large volumes of cancer genomic profiles. First we show that our approach successfully captures known SL-partners of tumor suppressors and oncogenes. Second, we construct a genome-wide network of SL-interactions in cancer and demonstrate its value in predicting gene essentiality. Third, we show that the under-expression of SL-partners has a strong positive prognostic value for cancer survival. Finally, we identify synthetic lethality arising from gene over-activation and show that it can be successfully used for predicting drug efficacy. These results form a computational basis for harnessing synthetic lethality to uncover cancer specific susceptibilities.
For more Information: