Geometry-constrained prediction of catalytically competent kinase domains across the human kinome

Accurate structural modeling of kinases across functional states helps to understand signaling networks. However, most human kinases lack experimentally determined active-state structures, and existing computational methods often fail to predict catalytically competent conformations. Here, we present a geometry-constrained modeling framework based on AFEXplorer (AFEX) that enables robust prediction of active, ATP- and magnesium-bound states for 436 catalytic kinase domains across the human kinome. By explicitly enforcing biophysical criteria, AFEX overcomes the template dependence and conformational bias inherent in AlphaFold2 and AlphaFold3. Benchmarking against 156 kinase-substrate complexes shows that AFEX achieves a sub-2 Å backbone root-mean-square deviation (RMSD) in 99% of cases and accurately recapitulates critical catalytic features with precise ATP coordination. Notably, AFEX correctly predicted active conformations for epidermal growth factor receptor (EGFR) and cyclin-dependent kinase 2 (CDK2), whereas AlphaFold3 failed. These improvements are consistent across kinase families, including both well-characterized kinases and those underrepresented in AlphaFold's training data. AFEX provides a generalizable framework for studying active conformations and dynamics in kinases.