ASTEROID: A Spatiotemporal Information Transformer for Forecasting Multi-Step Time Series of Molecular Dynamics

2026-06-16

AI Summary

This work addresses the high computational cost and inefficiency of long-timescale, multi-step prediction in molecular dynamics simulations by proposing an end-to-end Transformer-based framework. The approach models molecular trajectories as high-dimensional spatiotemporal sequences and directly predicts multi-step atomic coordinates, circumventing conventional iterative integration. It innovatively integrates local-global self-attention mechanisms within an encoder-decoder architecture and embeds spatiotemporal transformation equations to effectively capture multiscale spatiotemporal dependencies inherent in molecular systems. Evaluated on multiple quantum-mechanical molecular datasets, the method significantly outperforms existing models, achieving higher multi-step prediction accuracy while substantially reducing computational overhead, thereby enabling efficient long-timescale simulations.

Abstract

Molecular dynamics (MD) simulation is computationally demanding, particularly for large-scale systems requiring long-term analysis. Accurate forecast of the outcomes of a MD simulation is not only an attractive scientific challenge but also has substantial practical value. In this work, we developed a data-driven framework, termed ASTEROID (Advanced Spatiotemporal TransformER fOr Inferring Dynamics), that can directly predict multi-step atomic coordinates, avoiding conventional iterative integration. For this purpose, our ASTEROID reformulates MD trajectories as high-dimensional spatiotemporal sequences and integrates the Spatiotemporal Information (STI) Transformation equation into a Transformer architecture. The core innovation of ASTEROID lies in its ability to model multiscale spatiotemporal dependencies. In particular, for spatial dependencies, a local-global self-attention mechanism captures both short- and long-range interactions. For temporal dependencies, an encoder-decoder structure integrates global context with autoregressive forecasting. ASTEROID was evaluated on several quantum-mechanics derived molecular datasets. Our results indicate that ASTEROID achieved not only a higher level of accuracy in multi-step prediction than existing methods on various benchmarks, but also significantly reduced computational cost of conventional MD simulation. Moreover, the model supports iterative multi-step forecasting over an extended time scale. This work establishes a robust and generalizable data-driven paradigm for accelerating MD simulations.

Problem

• molecular dynamics
• time series forecasting
• spatiotemporal dependencies
• multi-step prediction
• computational cost

Innovation

• Spatiotemporal Transformer
• Molecular Dynamics Forecasting
• Multi-step Prediction
• Local-Global Self-Attention
• Data-driven Simulation