Prospective Evaluation of Structure-Based Simulations Reveal Their Ability to Predict the Impact of Kinase Mutations on Inhibitor Binding

• DOI: 10.1021/acs.jpcb.4c07794
• Publication Date: 2025-03-07T19:23:36Z
• Authors: Sukrit Singh; Vytautas Gapsys; Matteo Aldeghi; David Schaller; Aziz M. Rangwala; Jessica B. White; Joseph P. Bluck; Jenke Scheen; William G. Glass; Jiaye Guo; Sikander Hayat; Bert L. de Groot; Andrea Volkamer; Clara D. Christ; Markus A. Seeliger; John D. Chodera

Small molecule kinase inhibitors are critical in the modern treatment of cancers, evidenced by existence over 80 FDA-approved small-molecule inhibitors. Unfortunately, intrinsic or acquired resistance, often causing therapy discontinuation, is frequently caused mutations therapeutic target. The advent clinical tumor sequencing has opened additional opportunities for precision oncology to improve patient outcomes pairing optimal therapies with mutation profiles. However, efforts hindered lack sufficient biochemical evidence classify each as resistant sensitive existing Structure-based methods show promising accuracy retrospective benchmarks at predicting whether a will perturb inhibitor binding, but comparisons made pooling disparate experimental measurements across different conditions. We present first prospective benchmark structure-based approaches on blinded dataset in-cell affinities Abl mutants using NanoBRET reporter assay. compare results and their ability estimate impact binding (measured ΔΔG). Comparing physics-based simulations, Rosetta, previous machine learning models, we find that accurately inhibitor-resistant inhibitor-sensitizing, approach similar degree accuracy. simulations best suited ΔΔG distal active site. To probe modes failure, retrospectively investigate two clinically significant poorly predicted our methods, T315A L298F, starting configurations protonation states significantly alter predictions. Our computational provide estimating affinity future models. These have potential utility identifying tumor-specific mutations, resistance absence data, sensitizing established