Computational correction of off-targeting for CRISPR-Cas9 essentiality screens

Abstract

<jats:p>Off-target cleavage by Cas9 can confound measurements of cell proliferation/viability in CRISPR assays by eliciting a DNA-damage response that includes cell cycle arrest<jats:sup>1-3</jats:sup>. This gene-independent toxicity has been documented in large scale assays<jats:sup>2-4</jats:sup> and shown to be a source of false-positives when libraries are populated by promiscuous guide RNAs (gRNAs)<jats:sup>7</jats:sup>. To address this, we developed CSC, a computational method to correct for the effect of specificity on gRNA depletion. We applied CSC to screening data from the Cancer Dependency Map and show that it significantly improves the specificity of CRISPR-Cas9 essentiality screens while preserving known gene essentialities even for genes targeted by highly pro-miscuous guides. We packaged CSC in a Python software to allow its seamless integration into current CRISPR analysis pipelines and improve the sensitivity of essentiality screens for repetitive genomic loci.</jats:p>