ChronoCast: A time-series paradigm for molecular dynamics simulation using equivariant graph neural networks

Physical Review E - Statistical, Nonlinear, and Soft Matter Physics. 2026. Vol. 113. Article 65305.

Кузькин В. А., Кривцов А. М., Wang Y., Zhang Z., Lu S., Shan S., Chen J.

We propose a time-series paradigm named ChronoCast for molecular dynamics simulation using an advanced autoregressive Equivariant Graph Neural Network, which reformulates the conventionally time-consuming integration process into a forecasting task. This reformulation in turn offers a promising solution to overcome the challenge of tremendous computational cost for studying complex physical problems based on traditional algorithms. By incorporating velocity as a node feature and enforcing momentum conservation, the proposed model achieves exceptionally high accuracy in reproducing diverse physical properties. Using the simple Si crystal and complex van der Waals NbSe3 nanowires as two examples, we demonstrate that the radial distribution function, mean-squared displacement, and vibrational density of states in both systems can be accurately reproduced by long-term autoregressive forecasting. More importantly, ChronoCast significantly reduces the trajectory generation time by orders of magnitude compared to the ab initio method and by more than half compared to state-of-the-art machine learning potential, demonstrating superior efficiency over these prevailing molecular dynamics simulations. This work offers an accurate and efficient time-series approach for studying the statistical physics of complex dynamical processes.

Keywords: прогнозирование временных рядов, Моделирование молекулярной динамики, Molecular dynamics simulation, time-series forecasting, Equivariant Graph Neural Network, Autoregressive paradigm, Эквивариантные графовые нейронные сети, Авторегрессионная парадигма

Publication based on the results of: Analytical and computer simulation of wave energy transfer at different length scales (2025)