Fig. 1

Schematic diagram of the work-flow. First, GDSC cancer cell line drug sensitivity data, CCLE cancer cell gene expression data and TCGA/GTEx tissue gene expression data are combined and transformed. The CCLE gene expression data and GDSC drug sensitivity data (collectively referred to as the cell-line data) were used to build predictive models that were subsequently used to predict/impute the tissue drug sensitivity for the TCGA and GTEx samples. Broadly, for each drug, we divided the cell-line data into a training and testing set. We aimed to identify a 30-gene set whose gene expression levels are most predictive of the IC₅₀ values of the drug for the samples in the testing set. The resulting model (a 30-gene set) was subsequently used to predict the IC₅₀ value of the TCGA/GTEx samples. This process was repeated 100 times independently. The predicted IC₅₀ values from the 100 runs were then averaged and taken as the predicted IC₅₀ value of the drug for the samples. For details, see Methods