Systematic dose-response profiling of 133 kinase inhibitors across five cancer cell lines, generating 17.1 million peptidoform curves, establishes potency coherence analysis as a unified framework that identifies 5,318 confident kinase-substrate relationships for 96 human kinases, critically re-evaluates existing annotations revealing high false-positive rates in public databases, and enables estimation of kinase activities in cancer patient biopsies that are independent of expression levels and reveal entity-specific signaling patterns with therapeutic implications.

Mass-spectrometry-based phosphoproteomics enables the analysis of thousands of protein phosphorylation events across the human proteome. However, there is a lack of scalable, hypothesis-free, and statistically sound approaches for discovering, evaluating, and falsifying kinase::substrate relationships (KSRs). Here, we developed a new concept termed potency-coherence analysis. By measuring and integrating 17 million peptidoform-specific dose-response curves for 133 kinase inhibitors with known targets and affinities, we could critically re-evaluate published KSRs and discover thousands of potency-coherent and motif-plausible new KSRs for 96 human kinases. Application of these high-confidence KSRs enabled the estimation of kinase and signaling pathway activities in cancer patient biopsies. This unified and extendable framework has been implemented in ProteomicsDB to aid researchers in understanding the human phosphoproteome in health and disease.