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

Ziyang Yu, Wenbing Huang, Yang Liu

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 **Uni**fied **Sim**ulator (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.

Lay Summary:

Simulating how molecules move and interact at the atomic level is crucial for designing medicines and materials, but existing methods face a dilemma: they’re either too slow for practical use or too simplified to capture real-world complexity. Recent AI advancements only work for specific molecule types, limiting their real-world applications.

We developed UniSim, a universal AI "molecular simulator" trained on diverse molecules—from simple chemicals to complex proteins. By learning patterns across different molecular families, UniSim predicts atomic interactions accurately while adapting quickly to new systems. Its design complies with the underlying principles of physics by mimicking "virtual" forces during simulation.

Unlike prior tools confined to narrow domains, UniSim achieves highly competitive performance across small molecules, peptides, and proteins in our experiments. This breakthrough could accelerate discoveries in drug development by enabling realistic, efficient simulations for previously incompatible molecular systems.

Application-Driven Machine Learning: This submission is on Application-Driven Machine Learning.

Link To Code: https://github.com/yaledeus/UniSim

Primary Area: Applications->Chemistry, Physics, and Earth Sciences

Keywords: molecular dynamics, unified 3D molecular pretraining, forward simulation, cross-domain transferability

Submission Number: 10064