Deep Coarse-grained Potentials via Relative Entropy Minimization

Abstract

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 at 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 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 accuracy and reliability.

arXiv commentThis is a post-peer-review, pre-copyedit version of an article published in The Journal of Chemical Physics

References in corpus (25)

• Neural Message Passing for Quantum Chemistry
• E(3)-Equivariant Graph Neural Networks for Data-Efficient and Accurate Interatomic Potentials
• Big Data meets Quantum Chemistry Approximations: The $\Delta$-Machine Learning Approach
• Less is more: sampling chemical space with active learning
• A Fourth-Generation High-Dimensional Neural Network Potential with Accurate Electrostatics Including Non-local Charge Transfer
• Active Learning of Uniformly Accurate Inter-atomic Potentials for Materials Simulation
• SchNet: A continuous-filter convolutional neural network for modeling quantum interactions
• MACE: Higher Order Equivariant Message Passing Neural Networks for Fast and Accurate Force Fields
• OrbNet: Deep Learning for Quantum Chemistry Using Symmetry-Adapted Atomic-Orbital Features
• TorchMD: A deep learning framework for molecular simulations
• Machine Learning for Quantum Mechanical Properties of Atoms in Molecules
• Coarse Graining Molecular Dynamics with Graph Neural Networks
• DeePCG: constructing coarse-grained models via deep neural networks
• Forces are not Enough: Benchmark and Critical Evaluation for Machine Learning Force Fields with Molecular Simulations
• Machine Learning in QM/MM Molecular Dynamics Simulations of Condensed-Phase Systems
• Silicon liquid structure and crystal nucleation from ab-initio deep Metadynamics
• Learning neural network potentials from experimental data via Differentiable Trajectory Reweighting
• The Design Space of E(3)-Equivariant Atom-Centered Interatomic Potentials
• Automated discovery of a robust interatomic potential for aluminum
• Ultra-fast interpretable machine-learning potentials
• Metadynamics for Training Neural Network Model Chemistries: a Competitive Assessment
• Machine Learning Implicit Solvation for Molecular Dynamics
• Understanding three-body contributions to coarse-grained force-fields
• Path-space variational inference for non-equilibrium coarse-grained systems
• Contrastive Learning of Coarse-Grained Force Fields

Cited by in corpus (4)

• chemtrain: Learning Deep Potential Models via Automatic Differentiation and Statistical Physics
• JaxSGMC: Modular stochastic gradient MCMC in JAX
• Energy-Based Coarse-Graining in Molecular Dynamics: A Flow-Based Framework without Data
• Morphology-Aware Peptide Discovery via Masked Conditional Generative Modeling