Abstract 1555: Pan-cancer patterns of synthetic lethality: statistical modeling of gene dependency profiles

We present a methodology to fit statistical models of cell-viability profiles from RNAi-gene knockdowns. This method allows us to classify genes according to the degree of skewness in their viability distributions. The set of genes with the highest degree of skewness is highly enriched with many known oncogenes and tumor suppressors. We characterize many of these genes, compare them against the results of large sequencing efforts, and use them as inputs to a matrix-decomposition procedure that identifies the most salient cell viabilities shared by different cancer types. We catalog these pan-cancer patterns of synthetic lethality and characterize them by the genomic, transcriptional, and phenotypic features. This analysis provides a rich catalog of the most salient Achilles’ Heels of Pan-Cancer that can be helpful to identify new therapeutic strategies across cancers. Citation Format: Huwate Yeerna, Ramya Rangan, Andrew Aguirre, William Kim, Francisca Vazquez, Barbara Weir, Mahmoud Ghandi, Aviad Tsherniak, Jesse Boehm, William Hahn, Jill Mesirov, Pablo Tamayo. Pan-cancer patterns of synthetic lethality: statistical modeling of gene dependency profiles [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2017; 2017 Apr 1-5; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2017;77(13 Suppl):Abstract nr 1555. doi:10.1158/1538-7445.AM2017-1555