Generating transition states of chemical reactions via distance-geometry-based flow matching
Preprint
Abstract
Transition states (TSs) are crucial for understanding reaction mechanisms, yet their exploration is limited by the complexity of experimental and computational approaches. Here we propose TS-DFM, a flow matching framework that predicts TSs from reactants and products. By operating in molecular distance geometry space, TS-DFM explicitly captures the dynamic changes of interatomic distances in chemical reactions. A network structure named TSDVNet is designed to learn the velocity field for generating TS geometries accurately. On the benchmark dataset Transition1X, TS-DFM outperforms the previous state-of-the-art method React-OT by 30\% in structural accuracy. These predicted TSs provide high-quality initial structures, accelerating the convergence of CI-NEB optimization. Additionally, TS-DFM can identify alternative reaction paths. In our experiments, even a more favorable TS with lower energy barrier is discovered. Further tests on RGD1 dataset confirm its strong generalization ability on unseen molecules and reaction types, highlighting its potential for facilitating reaction exploration.Keywords
All Related Versions
This publication has 36 references indexed in Scilit:
- Machine Learning Transition State Geometries and Applications in Reaction Property PredictionWIREs Computational Molecular Science, 2025
- Sella, an Open-Source Automation-Friendly Molecular Saddle Point OptimizerJournal of Chemical Theory and Computation, 2022
- Hessian Matrix Update Scheme for Transition State Search Based on Gaussian Process RegressionJournal of Chemical Theory and Computation, 2020
- PhysNet: A Neural Network for Predicting Energies, Forces, Dipole Moments, and Partial ChargesJournal of Chemical Theory and Computation, 2019
- Computational Transition-State Design Provides Experimentally Verified Cr(P,N) Catalysts for Control of Ethylene Trimerization and TetramerizationACS Catalysis, 2018
- Methods for exploring reaction space in molecular systemsWIREs Computational Molecular Science, 2017
- Universal Structure Conversion Method for Organic Molecules: From Atomic Connectivity to Three‐Dimensional GeometryBulletin of the Korean Chemical Society, 2015
- A Unified Perspective on the Hydrogen Atom Transfer and Proton-Coupled Electron Transfer Mechanisms in Terms of Topographic Features of the Ground and Excited Potential Energy Surfaces As Exemplified by the Reaction between Phenol and RadicalsJournal of the American Chemical Society, 2008
- Prediction of Organic Reaction Products: Determining the Best Reaction ConditionsJournal of Chemical Information and Computer Sciences, 2000
- Shapes of simple polyatomic molecules and ions. I. Series HAAH and BAABJournal of the American Chemical Society, 1970