Reaction-Transformation-Aware Flow Matching for Generalizable Transition State Generation

2026-08-14

AI Summary

This study addresses the high computational cost of traditional transition state searches and the limited generalizability of existing machine learning approaches by proposing TransTS. This framework explicitly models atomic-level structural transformations between reaction endpoints to construct a unified geometric representation, integrating flow matching with equivariant neural networks for reaction-aware transition state generation. Experimental results demonstrate that TransTS significantly improves both transition state initialization quality and convergence rates across in-distribution and zero-shot out-of-distribution benchmarks. The method exhibits strong generalization to unseen reaction distributions, providing high-quality initial guesses for quantum chemistry calculations and effectively accelerating reaction pathway exploration.

Abstract

Transition-state (TS) structures define the energetic barriers and mechanistic pathways of elementary chemical reactions, yet their identification remains computationally demanding because conventional saddle-point searches require expensive quantum-mechanical calculations. Recent machine-learning approaches have accelerated TS generation by predicting structures from reaction endpoint information, but they primarily learn geometric correspondence between endpoints and TSs, leaving the structural transformations underlying elementary reactions implicitly represented. To address this limitation, we introduce TransTS, a reaction-transformation-aware framework for generalizable TS generation from atom-mapped reactant-product pairs. TransTS explicitly learns atom-level structural transformations between reaction endpoints and integrates them with a unified atom-aligned geometric representation of reactants, TSs and products, enabling reaction-aware equivariant generation of TS geometries. TransTS is designed to provide reliable TS initial guesses for subsequent quantum-chemical refinement, where generated structures are evaluated not only by geometric similarity but also by their ability to converge to validated saddle points and recover the intended reaction pathways. Across IID and zero-shot OOD benchmarks, TransTS demonstrates improved TS initialization quality, with particularly strong generalization to unseen reaction distributions. On the challenging GDB-10-rxn and GDB-17-rxn OOD benchmarks, TransTS generates TS candidates that more frequently converge to validated saddle points and recover the intended elementary reactions after refinement than existing approaches under the same training regime. Scaling reaction coverage and model capacity further improves both geometric fidelity and refinement outcomes.

Problem

• Transition State Generation
• Structural Transformation
• Generalizability
• Flow Matching

Innovation

• Transition State Generation
• Flow Matching
• Reaction-Transformation-Aware
• Equivariant Generation
• Out-of-Distribution Generalization