A phosphoproteome atlas of human cell lines reveals the landscape of kinase activity.

Protein phosphorylation orchestrates cellular signaling and controls most biological processes, with its dysregulation driving diseases, notably cancer. Comprehensive, high-throughput phosphoproteomics remains limited by detection sensitivity, data completeness and computational bottlenecks, especially in low-input settings. Here we present a comprehensive empirical human phosphoproteome resource, regrouping over 200,000 class I phosphosites across 33 diverse human cell lines. We demonstrate that this spectral library dramatically improves single-shot phosphoproteomics with 30-fold faster data processing compared with library-free approaches and enhances confidence in phosphosite localization even from minimal sample input. Integrating proteome and phosphoproteome data, we develop a combined kinase activity score (Cscore), revealing cell line- and cancer-specific signaling vulnerabilities, many correlating with drug sensitivity. This resource accelerates deep and reproducible phosphoproteomics, enables the systematic mapping of cellular signaling networks and may empower precision oncology by highlighting actionable kinase targets in diverse cell states.