Abstract
Cancer genomes often harbor hundreds of molecular aberrations. Such genetic variants can be drivers or passengers of tumorigenesis and create new vulnerabilities for potential therapeutic exploitation. To identify genotype-dependent vulnerabilities, a large number of forward genetic screens in different genetic backgrounds have been conducted. To aggregate and mine functional data from published screens in a standardized format, we recently developed GenomeCRISPR, a curated database for high-throughput CRISPR-Cas9 experiments. We then devised MINGLE, a computational framework that integrates genetic screens in large panels of cancer cell lines to predict genetic interactions between variant and knockout alleles. We showed that our method identifies specific known and previously unknown genetic relationships. For example, we identified a synthetic lethal dependency between RNF43 and members of the Glucosidase II complex. Additional experiments confirmed that the Glucosidase II complex acts as a regulator of Wnt/beta-catenin signalling. In our first study we analyzed a dataset consisting of screens performed in sixty cell lines. Since then the amount of data available for this analysis has increased several fold. We are continuously integrating the new data that become available as we work towards the next iteration of a genetic interaction map in cancer cells. We are specifically interested in using the increased resolution of this new map to understand how genetic networks differ between individual cancer types.
Citation Format: Benedikt Rauscher. Prediction of selective genetic vulnerabilities from large-scale functional screens in cancer cells [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2019; 2019 Mar 29-Apr 3; Atlanta, GA. Philadelphia (PA): AACR; Cancer Res 2019;79(13 Suppl):Abstract nr 2457.