---

基于CTRP数据库，以CELL数据集预测验证验证集药物的敏感性
参考文章：Chen Y , Xiaowen H , Yan L , et al. Prognosis and personalized treatment prediction in TP53-mutant hepatocellular carcinoma: an in silico strategy towards precision oncology[J]. Briefings in Bioinformatics. 2020, 1–13
Expression profile data and somatic mutation data of human cancer cell lines (CCLs) were obtained from the Broad Institute-Cancer Cell Line Encyclopedia (CCLE) project (https://portals.broadinstitute.org/ccle/) . The CERES scores of genome-scale CRISPR knockout screens for 18,333 genes in 739 cell lines were acquired from the dependency map (DepMap) portal (https://depmap.org/portal/). CERES score is used to measure the dependency of the gene of interest in certain CCL, and a lower score indicates a higher likelihood that the gene is essential in cell growth and survival of a given CCL.
Drug sensitivity data of CCLs were achieved from the Cancer Therapeutics Response Portal (CTRP v.2.0, released October 2015, https://portals.broadinstitute.org/ctrp) and PRISM Repurposing dataset (19Q4, released December 2019, https://depmap.org/portal/prism/). The CTRP contains the sensitivity data for 481 compounds over 835 CCLs and the PRISM contains the sensitivity data for 1448 compounds over 482 CCLs. Both two datasets provide the area under the dose–response curve (area under the curve—AUC) values as a measure of drug sensitivity, and lower AUC values indicate increased sensitivity to treatment. K-nearest neighbor (k-NN) imputation was applied to impute the missing AUC values. Before imputation, compounds with more than 20% of missing data were excluded. Because the CCLs in both datasets were obtained from the CCLE project, molecular data in CCLE were thus used for subsequent CTRP and PRISM analyses.