Transferable Learning of Reaction Pathways from Geometric Priors

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Nam, Juno, Steiner, Miguel, Misterka, Max, Yang, Soojung, Singhal, Avni, Gómez-Bombarelli, Rafael
Formato: Preprint
Publicado: 2025
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866908837919850496
author Nam, Juno
Steiner, Miguel
Misterka, Max
Yang, Soojung
Singhal, Avni
Gómez-Bombarelli, Rafael
author_facet Nam, Juno
Steiner, Miguel
Misterka, Max
Yang, Soojung
Singhal, Avni
Gómez-Bombarelli, Rafael
contents Identifying minimum-energy paths (MEPs) is crucial for understanding chemical reaction mechanisms but remains computationally demanding. We introduce MEPIN, a scalable machine-learning method for efficiently predicting MEPs from reactant and product configurations, without relying on transition-state geometries or pre-optimized reaction paths during training. The task is defined as predicting deviations from geometric interpolations along reaction coordinates. We address this task with a continuous reaction path model based on a symmetry-broken equivariant neural network that generates a flexible number of intermediate structures. The model is trained using an energy-based objective, with efficiency enhanced by incorporating geometric priors from geodesic interpolation as initial interpolations or pre-training objectives. Our approach generalizes across diverse chemical reactions and achieves accurate alignment with reference intrinsic reaction coordinates, as demonstrated on various small molecule reactions and [3+2] cycloadditions. Our method enables the exploration of large chemical reaction spaces with efficient, data-driven predictions of reaction pathways.
format Preprint
id arxiv_https___arxiv_org_abs_2504_15370
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Transferable Learning of Reaction Pathways from Geometric Priors
Nam, Juno
Steiner, Miguel
Misterka, Max
Yang, Soojung
Singhal, Avni
Gómez-Bombarelli, Rafael
Chemical Physics
Materials Science
Machine Learning
Identifying minimum-energy paths (MEPs) is crucial for understanding chemical reaction mechanisms but remains computationally demanding. We introduce MEPIN, a scalable machine-learning method for efficiently predicting MEPs from reactant and product configurations, without relying on transition-state geometries or pre-optimized reaction paths during training. The task is defined as predicting deviations from geometric interpolations along reaction coordinates. We address this task with a continuous reaction path model based on a symmetry-broken equivariant neural network that generates a flexible number of intermediate structures. The model is trained using an energy-based objective, with efficiency enhanced by incorporating geometric priors from geodesic interpolation as initial interpolations or pre-training objectives. Our approach generalizes across diverse chemical reactions and achieves accurate alignment with reference intrinsic reaction coordinates, as demonstrated on various small molecule reactions and [3+2] cycloadditions. Our method enables the exploration of large chemical reaction spaces with efficient, data-driven predictions of reaction pathways.
title Transferable Learning of Reaction Pathways from Geometric Priors
topic Chemical Physics
Materials Science
Machine Learning
url https://arxiv.org/abs/2504.15370