biostudies-literatureDarouich SStuttgart Center for Simulation Science, Universit?t StuttgartDeutsche ForschungsgemeinschaftState of Baden-W?rttembergMinisterium f?r Wissenschaft, Forschung und Kunst Baden-W?rttemberg2154-2165https://www.ebi.ac.uk/biostudies/studies/S-EPMC12933717biostudies-literatureUnknown66(4)Identifying transition states (TSs), the high-energy configurations that molecules pass through during chemical reactions, is essential for understanding and designing chemical processes. However, accurately and efficiently identifying these states remains one of the most challenging problems in computational chemistry. In this work, we introduce a new generative AI approach that improves the quality of initial guesses for TS structures. Our method can be combined with a variety of existing techniques, including both machine-learning models and fast, approximate quantum methods, to refine their predictions and bring them closer to chemically accurate results. Applied to TS guesses from a state-of-the-art machine-learning model, our approach reduces the median structural error to 0.077 Å anJournal of chemical information and modelingAdaptive Transition-State Refinement with Learned Equilibrium Flows.PMC12933717EXC2075 - 390740016INST 40/575-1 FUGGEXC 2120/1 - 390831618Darouich SNiepert MBien TKastner JTong VfalseAdaptive Transition-State Refinement with Learned Equilibrium Flows.Identifying transition states (TSs), the high-energy configurations that molecules pass through during chemical reactions, is essential for understanding and designing chemical processes. However, accurately and efficiently identifying these states remains one of the most challenging problems in computational chemistry. In this work, we introduce a new generative AI approach that improves the quality of initial guesses for TS structures. Our method can be combined with a variety of existing techniques, including both machine-learning models and fast, approximate quantum methods, to refine their predictions and bring them closer to chemically accurate results. Applied to TS guesses from a state-of-the-art machine-learning model, our approach reduces the median structural error to 0.077 Å an2026-01-01T00:00:00Z2026 Feb2026-07-16T20:47:11.136Z2026-07-10T03:12:25.898ZS-EPMC129337174162789810.1021/acs.jcim.5c02902