UniSim: A Unified Simulator for Time-Coarsened Dynamics of Biomolecules

Ziyang Yu, Wenbing Huang, Yang Liu
Proceedings of the 42nd International Conference on Machine Learning, PMLR 267:73097-73114, 2025.

Abstract

Molecular Dynamics (MD) simulations are essential for understanding the atomic-level behavior of molecular systems, giving insights into their transitions and interactions. However, classical MD techniques are limited by the trade-off between accuracy and efficiency, while recent deep learning-based improvements have mostly focused on single-domain molecules, lacking transferability to unfamiliar molecular systems. Therefore, we propose Unified Simulator (UniSim), which leverages cross-domain knowledge to enhance the understanding of atomic interactions. First, we employ a multi-head pretraining approach to learn a unified atomic representation model from a large and diverse set of molecular data. Then, based on the stochastic interpolant framework, we learn the state transition patterns over long timesteps from MD trajectories, and introduce a force guidance module for rapidly adapting to different chemical environments. Our experiments demonstrate that UniSim achieves highly competitive performance across small molecules, peptides, and proteins.

Cite this Paper

BibTeX
@InProceedings{pmlr-v267-yu25j, title = {{U}ni{S}im: A Unified Simulator for Time-Coarsened Dynamics of Biomolecules}, author = {Yu, Ziyang and Huang, Wenbing and Liu, Yang}, booktitle = {Proceedings of the 42nd International Conference on Machine Learning}, pages = {73097--73114}, year = {2025}, editor = {Singh, Aarti and Fazel, Maryam and Hsu, Daniel and Lacoste-Julien, Simon and Berkenkamp, Felix and Maharaj, Tegan and Wagstaff, Kiri and Zhu, Jerry}, volume = {267}, series = {Proceedings of Machine Learning Research}, month = {13--19 Jul}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v267/main/assets/yu25j/yu25j.pdf}, url = {https://proceedings.mlr.press/v267/yu25j.html}, abstract = {Molecular Dynamics (MD) simulations are essential for understanding the atomic-level behavior of molecular systems, giving insights into their transitions and interactions. However, classical MD techniques are limited by the trade-off between accuracy and efficiency, while recent deep learning-based improvements have mostly focused on single-domain molecules, lacking transferability to unfamiliar molecular systems. Therefore, we propose Unified Simulator (UniSim), which leverages cross-domain knowledge to enhance the understanding of atomic interactions. First, we employ a multi-head pretraining approach to learn a unified atomic representation model from a large and diverse set of molecular data. Then, based on the stochastic interpolant framework, we learn the state transition patterns over long timesteps from MD trajectories, and introduce a force guidance module for rapidly adapting to different chemical environments. Our experiments demonstrate that UniSim achieves highly competitive performance across small molecules, peptides, and proteins.} }
Endnote
%0 Conference Paper %T UniSim: A Unified Simulator for Time-Coarsened Dynamics of Biomolecules %A Ziyang Yu %A Wenbing Huang %A Yang Liu %B Proceedings of the 42nd International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2025 %E Aarti Singh %E Maryam Fazel %E Daniel Hsu %E Simon Lacoste-Julien %E Felix Berkenkamp %E Tegan Maharaj %E Kiri Wagstaff %E Jerry Zhu %F pmlr-v267-yu25j %I PMLR %P 73097--73114 %U https://proceedings.mlr.press/v267/yu25j.html %V 267 %X Molecular Dynamics (MD) simulations are essential for understanding the atomic-level behavior of molecular systems, giving insights into their transitions and interactions. However, classical MD techniques are limited by the trade-off between accuracy and efficiency, while recent deep learning-based improvements have mostly focused on single-domain molecules, lacking transferability to unfamiliar molecular systems. Therefore, we propose Unified Simulator (UniSim), which leverages cross-domain knowledge to enhance the understanding of atomic interactions. First, we employ a multi-head pretraining approach to learn a unified atomic representation model from a large and diverse set of molecular data. Then, based on the stochastic interpolant framework, we learn the state transition patterns over long timesteps from MD trajectories, and introduce a force guidance module for rapidly adapting to different chemical environments. Our experiments demonstrate that UniSim achieves highly competitive performance across small molecules, peptides, and proteins.
APA
Yu, Z., Huang, W. & Liu, Y.. (2025). UniSim: A Unified Simulator for Time-Coarsened Dynamics of Biomolecules. Proceedings of the 42nd International Conference on Machine Learning, in Proceedings of Machine Learning Research 267:73097-73114 Available from https://proceedings.mlr.press/v267/yu25j.html.

Related Material