Search papers, labs, and topics across Lattice.
This paper introduces ASTEROID, a novel framework that leverages a Transformer architecture to predict multi-step atomic coordinates in molecular dynamics simulations, bypassing traditional iterative integration methods. By reformulating molecular trajectories into high-dimensional spatiotemporal sequences and employing a local-global self-attention mechanism, ASTEROID effectively captures both short- and long-range spatial interactions, as well as temporal dependencies through an encoder-decoder structure. The results demonstrate that ASTEROID not only outperforms existing prediction methods in accuracy but also significantly reduces the computational costs associated with conventional molecular dynamics simulations.
ASTEROID achieves superior accuracy in multi-step predictions while slashing the computational costs of molecular dynamics simulations.
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.
