Despite significant progress in cancer research, current standard-of-care drugs fail to cure many types of cancers. Hence, there is an urgent need to identify better predictive biomarkers and treatment regimes (Jiang et al., Annual Review 2018). I developed CARE (Computational Analysis of REsistance), a statistical method for identification of genes associated with targeted therapy efficacy from compound screens (Jiang et al., Cell Systems 2018). CARE applies multivariate regressions to test how drug targets interact with other genes to affect drug response. Such interaction effects were not considered in previous studies but are critical in identifying biomarkers for targeted therapy response. When evaluated using clinical datasets, CARE can predict the therapy outcome better than signatures from other methods. CARE also identified PRKD3 as a regulator of Lapatinib resistance. Experimental validation by both siRNA and compounds confirmed that inhibition of PRKD3 significantly sensitizes HER2+ breast cancer cells to both Lapatinib and Trastuzumab.

• Graphical Abstract of CARE (website URL here: CARE website)

• Graphical Abstract from our Cancer Data Science Review (Jiang et al., Annual Review 2018)