Thaler Stephan, Stupp Maximilian, Zavadlav Julija
Multiscale Modeling of Fluid Materials, Department of Engineering Physics and Computation, TUM School of Engineering and Design, Technical University of Munich, Munich, Germany.
J Chem Phys. 2022 Dec 28;157(24):244103. doi: 10.1063/5.0124538.
Neural network (NN) potentials are a natural choice for coarse-grained (CG) models. Their many-body capacity allows highly accurate approximations of the potential of mean force, promising CG simulations of unprecedented accuracy. CG NN potentials trained bottom-up via force matching (FM), however, suffer from finite data effects: They rely on prior potentials for physically sound predictions outside the training data domain, and the corresponding free energy surface is sensitive to errors in the transition regions. The standard alternative to FM for classical potentials is relative entropy (RE) minimization, which has not yet been applied to NN potentials. In this work, we demonstrate, for benchmark problems of liquid water and alanine dipeptide, that RE training is more data efficient, due to accessing the CG distribution during training, resulting in improved free energy surfaces and reduced sensitivity to prior potentials. In addition, RE learns to correct time integration errors, allowing larger time steps in CG molecular dynamics simulation, while maintaining accuracy. Thus, our findings support the use of training objectives beyond FM, as a promising direction for improving CG NN potential's accuracy and reliability.
神经网络（NN）势是粗粒度（CG）模型的自然选择。它们的多体能力允许对平均力势进行高度精确的近似，有望实现前所未有的高精度CG模拟。然而，通过力匹配（FM）自下而上训练的CG NN势存在有限数据效应：它们依赖于先验势在训练数据域之外进行合理的物理预测，并且相应的自由能面对过渡区域中的误差很敏感。对于经典势，FM的标准替代方法是相对熵（RE）最小化，该方法尚未应用于NN势。在这项工作中，我们针对液态水和丙氨酸二肽的基准问题证明，由于在训练期间访问了CG分布，RE训练在数据效率上更高，从而得到改进的自由能面并降低了对先验势的敏感性。此外，RE学会校正时间积分误差，在保持精度的同时允许在CG分子动力学模拟中采用更大的时间步长。因此，我们的研究结果支持使用除FM之外的训练目标，作为提高CG NN势的准确性和可靠性的一个有前景的方向。
