MAPLE: A General Framework for Automated Molecular Modeling across Machine-Learning Potentials

Machine-learning force fields (MLFFs) are reshaping computational chemistry and biology by providing near–quantum mechanical accuracy at a computational cost comparable to conventional force fields, enabling applications in biomolecular simulation, catalysis, and materials science. However, toward broader application of MLFFs, a unified and automated computational platform remains lacking. Here, we develop MAPLE, a program specifically designed for MLFF-based molecular modeling, featuring a tailored framework and parallelized algorithms. In this study, we demonstrate the robustness and usability of MAPLE through systematic benchmarking of modern reactive MLFFs and applications to multiple biocatalytic scenarios, highlighting its capability for fast yet accurate simulation of catalytic reactions. By leveraging the accuracy and efficiency of MLFFs together with a highly optimized and flexible software framework, MAPLE serves as a next-generation, physically informed, machine-learning–driven molecular modeling platform with broad applicability to rational catalyst design and drug discovery.