Abstract:We develop a new deep potential—range correction (DPRc) machine learning potential for combined quantum mechanical/molecular mechanical (QM/MM) simulations of chemical reactions in the condensed phase. The new range correction enables short-ranged QM/MM interactions to be tuned for higher accuracy, and the correction smoothly vanishes within a specified cutoff. We further develop an active learning procedure for robust neural network training. We test the DPRc model and training procedure against a series of six nonenzymatic phosphoryl transfer reactions in solution that are important in mechanistic studies of RNA-cleaving enzymes. Specifically, we apply DPRc corrections to a base QM model and test its ability to reproduce free-energy profiles generated from a target QM model. We perform these comparisons using the MNDO/d and DFTB2 semiempirical models because they differ in the way they treat orbital orthogonalization and electrostatics and produce free-energy profiles which differ significantly from each other, thereby providing us a rigorous stress test for the DPRc model and training procedure. The comparisons show that accurate reproduction of the free-energy profiles requires correction of the QM/MM interactions out to 6 Å. We further find that the model's initial training benefits from generating data from temperature replica exchange simulations and including high-temperature configurations into the fitting procedure, so the resulting models are trained to properly avoid high-energy regions. A single DPRc model was trained to reproduce four different reactions and yielded good agreement with the free-energy profiles made from the target QM/MM simulations. The DPRc model was further demonstrated to be transferable to 2D free-energy surfaces and 1D free-energy profiles that were not explicitly considered in the training. Examination of the computational performance of the DPRc model showed that it was fairly slow when run on CPUs but was sped up almost 100-fold when using NVIDIA V100 GPUs, resulting in almost negligible overhead. The new DPRc model and training procedure provide a potentially powerful new tool for the creation of next-generation QM/MM potentials for a wide spectrum of free-energy applications ranging from drug discovery to enzyme design.The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acs.jctc.1c00201.Training errors (energy unit: eV, force unit: eV/Å) for the native reaction with different ML parameter sets; ratio of frames with different model deviations in the final simulations for the native reaction with different ML parameter sets; comparison of target (DFTB2) and model (MNDO/d+ML) transition-state coordinate values; transition-state (TS) barrier height and reaction coordinate values for the 2D study of the native reaction, with DFTB2 and MNDO/d Hamiltonians and ML models trained only on 1D reaction; training errors and histogram of frame energies; forces in the QM region; forces in the MM region; free-energy profiles of different variants, compared with original DFTB2 and MNDO/d curves; and illustration of scaling of the wall clock time per simulation step (ms/step) observed in the DFTB2 QM/MM simulations of the native reaction with and without the use of DPRc corrections CPU and GPU timings plotted separately(PDF)This article has not yet been cited by other publications.

MLRNet: Combining the Physics-Motivated Potential Models with Neural Networks for Intermolecular Potential Energy Surface Construction.

Global Neural Network Potential with Explicit Many-Body Functions for Improved Descriptions of Complex Potential Energy Surface.

A neural network potential based on pairwise resolved atomic forces and energies

AIMNet2: A Neural Network Potential to Meet your Neutral, Charged, Organic, and Elemental-Organic Needs

LAMMPS Implementation of Rapid Artificial Neural Network Derived Interatomic Potentials

Neural Network Potential with Multi-Resolution Approach Enables Accurate Prediction of Reaction Free Energies in Solution

Extending the atomic decomposition and many-body representation, a chemistry-motivated monomer-centered approach for machine learning potentials

Machine Learning Interatomic Potentials and Long-Range Physics

Modeling Intermolecular Interactions With XDM Dispersion Corrections to Neural Network Potentials

Spline-based neural network interatomic potentials: blending classical and machine learning models

Heterogeneous Relational Message Passing Networks for Molecular Dynamics Simulations

Learning Interatomic Potentials at Multiple Scales

PhysNet Meets CHARMM: A Framework for Routine Machine Learning / Molecular Mechanics Simulations

Transferable Water Potentials Using Equivariant Neural Networks

A deep potential model with long-range electrostatic interactions

Highly Accurate Description of Long-Range Interactions through the Combination of Neural Networks and Physical Models

Including physics-informed atomization constraints in neural networks for reactive chemistry

Development of Range-Corrected Deep Learning Potentials for Fast, Accurate Quantum Mechanical/Molecular Mechanical Simulations of Chemical Reactions in Solution

Permutationally Invariant Fitting of Intermolecular Potential Energy Surfaces: A Case Study of the Ne-C2H2 System.

Ab initio Accuracy Neural Network Potential for Drug-like Molecules

Challenges for Machine Learning Force Fields in Reproducing Potential Energy Surfaces of Flexible Molecules