跨晶体对称性的分子动力学可迁移图神经网络代理模型

中文摘要 AI 辅助

我们提出了一种用于分子动力学（MD）的可迁移图神经网络（GNN）代理框架，该框架直接预测原子位移并传播原子构型，无需显式的力评估或数值时间积分。本工作的核心目标是确定一个通用的GNN公式能否表示具有根本不同晶体对称性和配位环境的材料中的原子动力学。我们将相同的网络架构、特征表示、图构建和训练协议应用于面心立方（FCC）铝、体心立方（BCC）铁和六方密堆积（HCP）镁，而无需针对对称性进行修改。在这些不同的元素和晶体系统中，该框架实现了约10^-4埃^2量级的位置预测误差，并支持稳定的自回归传播至纳秒时间尺度。预测的轨迹保持了热力学稳定性、径向分布函数中特征性的配位壳层结构以及随温度变化的均方位移行为。特别是，该框架无需引入晶格特定的表示即可捕捉BCC铁的紧密间隔配位壳层和HCP镁的各向异性配位环境。这些结果表明，基于GNN的直接原子传播可以被构建为一种跨不同元素、晶格对称性和配位几何的可迁移框架，为加速分子动力学的通用代理模型提供了一条途径。

英文摘要

We present a transferable graph neural network (GNN) surrogate framework for molecular dynamics (MD) that directly predicts atomic displacements and propagates atomistic configurations without explicit force evaluation or numerical time integration. The central objective of this work is to establish whether a common GNN formulation can represent atomic dynamics across materials with fundamentally different crystal symmetries and coordination environments. The same network architecture, feature representation, graph construction, and training protocol are applied without symmetry-specific modification to face-centered cubic (FCC) aluminum, body-centered cubic (BCC) iron, and hexagonal close-packed (HCP) magnesium. Across these distinct elemental and crystallographic systems, the framework achieves position-prediction errors on the order of 10^-4 Angstrom^2 and supports stable autoregressive propagation to nanosecond time scales. The predicted trajectories preserve thermodynamic stability, characteristic coordination-shell structure in the radial distribution functions, and temperature-dependent mean-squared-displacement behavior. In particular, the framework captures the closely spaced coordination shells of BCC iron and the anisotropic coordination environment of HCP magnesium without introducing lattice-specific representations. These results demonstrate that direct GNN-based atomic propagation can be formulated as a transferable framework across different elements, lattice symmetries, and coordination geometries, providing a pathway toward generalizable surrogate models for accelerated molecular dynamics.

发表机构

• Merrimack College（梅里马克学院）
• West Virginia State University（西弗吉尼亚州立大学）

机构由 AI 辅助整理，请以论文原文为准。