(197y) Pysages: Enhanced Sampling for Ab Initio Dynamics and Machine Learning Potentials

PySAGES is a user-friendly Python library that enables enhanced sampling in molecular dynamics simulations by providing a simple interface to write and leverage complex collective variables and enhanced sampling methods. In this work, we demonstrate the coupling of PySAGES with Ab Initio integrators using the ASE interface, which facilitates enhanced sampling in systems that require first-principles accuracy. Additionally, we showcase the use of PySAGES as a tool for studying the robustness of machine-learned force fields by evaluating different examples of DeePMD, GAP, and Graph Neural Network potentials and assessing their reliability in preserving the behavior of a system in terms of selected molecular descriptors.

Our work highlights the versatility of PySAGES and its potential to accelerate the study and analysis of molecular dynamics simulations, as well as the design of molecules and materials in various fields.