Generating transition states of chemical reactions via distance-geometry-based flow matching

2025-11-21

AI Summary

Predicting transition states (TS) of chemical reactions remains challenging due to the high computational and experimental costs of exploration. To address this, we propose TS-DFM: a flow-matching framework operating in molecular distance geometry space, coupled with a novel TSDVNet architecture that explicitly models the dynamic evolution of interatomic distances to directly generate high-quality TS structures from reactants and products. Unlike conventional TS search methods reliant on initial reaction-path guesses, TS-DFM bypasses such constraints, enabling discovery of previously unknown reaction pathways and identification of lower-energy TS candidates. On the Transition1X benchmark, TS-DFM improves TS prediction accuracy by 30% over React-OT; generated TS structures significantly accelerate CI-NEB convergence; and on the cross-reaction-type RGD1 test set, TS-DFM demonstrates strong generalization. This work establishes an efficient, robust paradigm for mechanistic reaction analysis.

Technology Category

• Machine Learning: Time-Series/Data Streams
• Search and Optimization: Learning to Search
• Planning, Routing, and Scheduling: Planning with Language Models

Application Category

• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

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.

Problem

• Predicting transition states from reactants and products in chemical reactions
• Overcoming limitations of complex experimental and computational transition state exploration
• Identifying alternative reaction paths and discovering lower energy barrier transitions

Innovation

• Flow matching framework predicts transition states
• Operates in molecular distance geometry space
• TSDVNet learns velocity field for geometries