BackgroundCancer drugs often kill cells independent of their putative targets, suggesting the limitation of existing drug target information. The lack of understanding of a drug’s mechanism of action may prevent biomarker identification and ultimately lead to attrition in clinical trials. Current experimental strategies, such as binding affinity assays provide limited coverage at the proteome scale. In this study, we explored whether the integration of loss-of-function genetic and drug sensitivity screening data could define a novel signature to better understand the mechanisms of action of drugs.MethodsLoss-of-function genetic screening data was collected from the DepMap database, while drug sensitivity data were collected from three extensive screening studies, namely CTRP (n = 545), GDSC (n = 198), and PRISM (n = 1448). An L1 penalized regression model using the gene essentiality features was constructed for each drug to predict its sensitivity on multiple cell lines. The optimized model coefficients were then considered as the gene essentiality signature of the drug. We compared the gene essentiality signature with structure-based fingerprints and the gene expression signature of cancer drugs in predictions of their known targets. Finally, we applied the gene essentiality signature to predict the novel targets for a panel of noncancer drugs with potential anticancer efficacy.ResultsWe showed that the gene essentiality signature can predict drug targets and their downstream signaling pathways. Both supervised and unsupervised prediction accuracies were higher than those using chemical fingerprints and gene expression signatures. Pathway analyses of these gene essentiality signatures confirmed key mechanisms previously reported, including the EGFR signaling network for lapatinib, and DNA mismatch repair drugs. Finally, we showed that the gene essentiality signature of noncancer drugs can discover novel targets.ConclusionsIntegrating drug sensitivity data and loss-of-function genetic data enables the construction of gene essentiality signatures that help discover drug targets and their downstream signaling pathways. We found novel targets for noncancer drugs that explain their anticancer efficacy, paving the way for the rational design of drug repurposing.