Development of a Deep Reinforcement Learning Approach for Transition State Search in Molecular Systems: A Transferable Model Across Structural and Dimensional Variations
Abstract
We developed a DRL approach for TS searches combining the deep Q-network and the bidirectional long short-term memory architectures to handle variable-sized molecular systems. We applied our method to 1,3,5-hexatriene (Triene), alanine dipeptide (Ala2), and 1,3,5,7-octatetraene (Tetraene) as target systems with the search space limited to backbone dihedral angles. Our model successfully learned the TS search tasks for all examined systems. More importantly, we demonstrate that our model pre-trained on Triene exhibits distinct transfer-learning capabilities across both backbone modifications (Triene to Ala2) and chain extensions (Triene to Tetraene). Such transferability is highly promising for future applications to molecules in general, as it allows an agent trained on simple systems to accelerate TS searches in larger or more complex ones. We expect that these findings on cross-system transferability represent an important step forward in developing DRL-based approaches for TS searches.
Related
- State-Constrained SHAP Distills Polymer Property Models into Transferable Structural Rules for Fixed-Interval Screening
- Efficient and Transferable Machine Learning Potentials for Zn-Coordinated Zeolitic Imidazolate Frameworks Through Structural Diversity Sampling and Transfer Learning.
- Development and Bayesian Uncertainty Quantification of Transferable Coarse-Grained Models for Carbohydrates
- Reinforcement learning amortizes transition-state physics into one-step flow models
- Effect of Graft Development in PLLA-Functionalized Cellulose Nanocrystals on Dispersibility, Transparency, and Mechanical Reinforcement in PLA Nanocomposites
BibTeX
@misc{kikkawa2026development,
title = {{Development of a Deep Reinforcement Learning Approach for Transition State Search in Molecular Systems: A Transferable Model Across Structural and Dimensional Variations}},
author = {Takechika Kikkawa and Shusuke Yamanaka and Takashi Kawakami and Mitsutaka Okumura},
howpublished = {arXiv},
year = {2026},
month = oct,
doi = {10.48550/arxiv.2610.39329},
eprint = {2610.39329},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2610.39329},
}
RIS
TY - UNPB
TI - Development of a Deep Reinforcement Learning Approach for Transition State Search in Molecular Systems: A Transferable Model Across Structural and Dimensional Variations
AU - Takechika Kikkawa
AU - Shusuke Yamanaka
AU - Takashi Kawakami
AU - Mitsutaka Okumura
PY - 2026
DA - 2026/10/29
PB - arXiv
DO - 10.48550/arxiv.2610.39329
UR - https://arxiv.org/abs/2610.39329
AB - We developed a DRL approach for TS searches combining the deep Q-network and the bidirectional long short-term memory architectures to handle variable-sized molecular systems. We applied our method to 1,3,5-hexatriene (Triene), alanine dipeptide (Ala2), and 1,3,5,7-octatetraene (Tetraene) as target systems with the search space limited to backbone dihedral angles. Our model successfully learned the TS search tasks for all examined systems. More importantly, we demonstrate that our model pre-trained on Triene exhibits distinct transfer-learning capabilities across both backbone modifications (Triene to Ala2) and chain extensions (Triene to Tetraene). Such transferability is highly promising for future applications to molecules in general, as it allows an agent trained on simple systems to accelerate TS searches in larger or more complex ones. We expect that these findings on cross-system transferability represent an important step forward in developing DRL-based approaches for TS searches.
ER -