Fig. 3: Evaluation of kinase-substrate libraries for kinase activity inference.

a Coverage of kinases across different kinase-substrate libraries including the number of unique kinases and kinases with at least five annotated targets in each library. b Coverage of kinase-substrate interactions across libraries and number of unique kinase-substrate interactions in each library. c Predictive performance of kinase-substrate libraries for kinase activity inference in identifying deregulated kinases from phosphoproteomics data using the perturbation-, activating site- and protein-based benchmarking approaches. For all libraries the z-score was used to infer kinase activities. The AUROC calculation was repeated a thousand times, with randomly selecting a subset of the negative classes with the same size as the positive class for the perturbation-based benchmark and 80% of the gold standard set of the tumor-based benchmark (n = 1000). d Evaluation of kinase activity inference for kinases with 5–10, 11–25 and >25 measured targets. Kinase activity was inferred using multiple kinase-substrate libraries in combination with the z-score method and evaluated using the perturbation-, activating site- and protein-based benchmarking approaches. For the perturbation-based approach the scaled rank was compared across kinases (n depicted in barplot) and the AUROC calculation was repeated a thousand times, with randomly selecting 80% of the gold standard set of the tumor-based benchmark (n = 1000). For the boxplots, the central line depicts the median, the box hinges represent the 25th to 75th percentiles, and the whiskers extend up to 1.5 times the interquartile range above and below the box hinges. Outliers are depicted as individual hollow points beyond the whiskers. Source data are provided as a Source Data file.