Basic stability tests of machine learning potentials for molecular simulations in computational drug discovery

Abstract

Neural network potentials trained on quantum-mechanical data can calculate molecular interactions with relatively high speed and accuracy. However, neural network potentials might exhibit instabilities, nonphysical behavior, or lack accuracy. To assess the reliability of neural network potentials, a series of tests is conducted during model training, in the gas phase, and in the condensed phase. The testing procedure is performed for eight in-house neural network potentials based on the ANI-2x dataset, using both the ANI-2x and MACE architectures. This allows an evaluation of the effect of the model architecture on its performance. We also perform stability tests of the publicly available neural network potentials ANI-2x, ANI-1ccx, MACE-OFF23, and AIMNet2. A normal mode analysis of 14 simple benchmark molecules revealed that the small MACE-OFF23 model shows large deviations from the reference quantum-mechanical energy surface. Also, some MACE models with a reduced number of parameters failed to produce stable molecular dynamics simulations in the gas phase, and all MACE models exhibit unfavorable behavior during steric clashes. The published ANI-2x and one of the in-house MACE models are not able to reproduce the structure of liquid water at ambient conditions, forming an amorphous solid phase instead. The ANI-1ccx model shows nonphysical additional energy minima in bond length and bond angle space, which caused a phase transition to an amorphous solid. Out of all 13 considered public and in-house models, only one in-house model based on the ANI-2x B97-3c dataset shows better agreement with the experimental radial distribution function of water than the simple molecular mechanics TIP3P model. This shows that great care must be taken during model training and when selecting a neural network potential for real-world applications.