Toward an integrated map of genetic interactions in cancer cells

Abstract

Cancer genomes often harbor hundreds of molecular aberrations. Such genetic variants can be drivers or passengers of tumorigenesis and create vulnerabilities for potential therapeutic exploitation. To identify genotype‐dependent vulnerabilities, forward genetic screens in different genetic backgrounds have been conducted. We devised MINGLE, a computational framework to integrate CRISPR/Cas9 screens originating from different libraries building on approaches pioneered for genetic network discovery in model organisms. We applied this method to integrate and analyze data from 85 CRISPR/Cas9 screens in human cancer cells combining functional data with information on genetic variants to explore more than 2.1 million gene‐background relationships. In addition to known dependencies, we identified new genotype‐specific vulnerabilities of cancer cells. Experimental validation of predicted vulnerabilities identified GANAB and PRKCSH as new positive regulators of Wnt/β‐catenin signaling. By clustering genes with similar genetic interaction profiles, we drew the largest genetic network in cancer cells to date. Our scalable approach highlights how diverse genetic screens can be integrated to systematically build informative maps of genetic interactions in cancer, which can grow dynamically as more data are included.

High‐resolution mapping of cancer cell networks using co‐functional interactions

“…The loadings on the first principal component, which again explained most of the variance in olfactory receptor scores, were highly correlated with the variance in effect size estimates for olfactory receptor genes for each cell line ( R = −0.89, Fig EV1C). One possibility is that regressing on the variance of each cell line acts as a technical correction, re‐centering and scaling the effect sizes in a manner similar to that performed in a similar study (Rauscher et al , ) and recommended in a recent article (McFarland et al , ). In this case, our scaling would roughly equalize variance in biological effect sizes among negative control genes.…”

“…The publication of hundreds of CRISPR screens in cell lines drawn from diverse cell lineage and mutational backgrounds has invited even broader surveys of co‐essentiality (Meyers et al , ; Data ref: Meyers et al , ). One study has dissected the composition of essential protein complexes (Pan et al , ), another has leveraged the natural occurrence of gene activating mutations to ascertain likely genetic interactions (Rauscher et al , ), and other work accessible as a preprint has focused on the organization of cancer growth pathways (preprint: Kim et al , ). In all these cases, interactions identified from correlated gene profiles operated on multiple levels of cellular regulation, validating parallel screening as a powerful tool for reconstructing cell networks.…”

KCML : a machine‐learning framework for inference of multi‐scale gene functions from genetic perturbation screens

“…We envision that systematic application of KCML to the large amounts of generated HT-GPS datasets will greatly accelerate and advance our understanding of gene functions at the molecular, cellular and tissue levels which can lead to the discovery of new therapeutic gene targets. (Szklarczyk et al, 2017) and https://www.pathwaycommons.org (Cerami et al, 2011) TCGA colorectal cancer data https://portal.gdc.cancer.gov (Muzny et al, 2012) Viability data (Rauscher et al, 2018) Expression data of MCF7 http://www.lincsproject.org/ (Duan et al, 2014) Sequence similarity of olfactory receptors https://genome.weizmann.ac.il/horde/ (Olender et al, 2013) Methods and Protocols KCML implementation and data analysis KCML pipeline and all analyses were performed using MATLAB (http://www.mathworks.com/) unless stated otherwise.…”

Biases and Blind-Spots in Genome-Wide CRISPR Knockout Screens

“…Extensive screening of cancer cell lines under the DepMap project-and, critically, the open availability of this data--affords an opportunity for re-evaluating the assumptions under which 415 these assays have been carried out. Notably, assumptions about replication and library-and batch-specific effects have been addressed in some detail Dempster et al, 2019;Rauscher et al, 2018), but questions about what might be systematically missing from these data have, to our knowledge, not been rigorously explored.…”