High-throughput knockout screens using CRISPR-Cas9 are now a widespread method for evaluating the essentiality of genes in different cell types. Here, we introduce a probabilistic model and inference framework, Analysis of CRISPR-based Essentiality (ACE), designed to test for differential signatures of essentiality between cell lines. ACE estimates the essentiality of each gene using a flexible likelihood framework based on the CRISPR-Cas9 experimental process and the observed sequencing

more »

... In addition, our method can identify which genes are essential only in a specified subset of samples by directly contrasting the likelihood of competing hypotheses - whether a gene has a constant or differential essentiality between samples. We show using simulations that our approach improves the accuracy of essentiality predictions compared to other methods, and is especially useful for the identification of weaker signals of essentiality. ACE performance was further validated on publicly available CRISPR screen data to distinguish between essential and nonessential genes. In summary, this method provides an improved quantification of essentiality specific to cancer subtypes, and a robust probabilistic framework to identify genes of interest.