Neural networks-based variationally enhanced sampling

Abstract

Sampling complex free energy surfaces is one of the main challenges of modern atomistic simulation methods. The presence of kinetic bottlenecks in such surfaces often renders a direct approach useless. A popular strategy is to identify a small number of key collective variables and to introduce a bias potential that is able to favor their fluctuations in order to accelerate sampling. Here we propose to use machine learning techniques in conjunction with the recent variationally enhanced sampling method [Valsson and Parrinello, Physical Review Letters 2014] to determine such potential. This is achieved by expressing the bias as a neural network. The parameters are determined in a reinforcement learning scheme aimed at minimizing the variationally enhanced sampling functional. This required the development of a new and more efficient minimization technique. The expressivity of neural networks allows accelerating sampling in systems with rapidly varying free energy surfaces, removing boundary effects artifacts, and making one more step towards being able to handle several collective variables.

References in corpus (21)

• Canonical sampling through velocity-rescaling
• Escaping free-energy minima
• PLUMED 2: New feathers for an old bird
• Well-Tempered Metadynamics: A Smoothly Converging and Tunable Free-Energy Method
• Gaussian Approximation Potentials: the accuracy of quantum mechanics, without the electrons
• Deep Potential Molecular Dynamics: a scalable model with the accuracy of quantum mechanics
• Solving the Quantum Many-Body Problem with Artificial Neural Networks
• Time-lagged autoencoders: Deep learning of slow collective variables for molecular kinetics
• A Variational Approach to Enhanced Sampling and Free Energy Calculations
• Metadynamics with adaptive Gaussians
• Non-strongly-convex smooth stochastic approximation with convergence rate O(1/n)
• Variational Encoding of Complex Dynamics
• Silicon liquid structure and crystal nucleation from ab-initio deep Metadynamics
• Reinforced dynamics for enhanced sampling in large atomic and molecular systems
• Learning Free Energy Landscapes Using Artificial Neural Networks
• Variationally Optimized Free Energy Flooding for Rate Calculation
• Making the best of a bad situation: a multiscale approach to free energy calculation
• Predictive Collective Variable Discovery with Deep Bayesian Models
• Multithermal-multibaric molecular simulations from a variational principle
• A variational approach to nucleation simulation
• Path collective variables without paths

Cited by in corpus (29)

• Machine learning for molecular simulation
• Combining Machine Learning and Computational Chemistry for Predictive Insights Into Chemical Systems
• Enhanced sampling methods for molecular dynamics simulations
• Data-Driven Collective Variables for Enhanced Sampling
• Machine learning force fields and coarse-grained variables in molecular dynamics: application to materials and biological systems
• A unified framework for machine learning collective variables for enhanced sampling simulations: $\text{mlcolvar}$
• Efficient sampling of high-dimensional free energy landscapes using adaptive reinforced dynamics
• A Perspective on Deep Learning for Molecular Modeling and Simulations
• Enhanced Sampling in the Age of Machine Learning: Algorithms and Applications
• Multiscale reweighted stochastic embedding (MRSE): Deep learning of collective variables for enhanced sampling
• Transferability of neural network potentials for varying stoichiometry: phonons and thermal conductivity of Mn${}_x$Ge${}_y$ compounds
• Unbiased Monte Carlo Cluster Updates with Autoregressive Neural Networks
• Generating valid Euclidean distance matrices
• Variational control forces for enhanced sampling of nonequilibrium molecular dynamics simulations
• Path sampling of recurrent neural networks by incorporating known physics
• Differentiable Molecular Simulations for Control and Learning
• Variational design principles for nonequilibrium colloidal assembly
• Collective Variables for Conformational Polymorphism in Molecular Crystals
• Deep learning the slow modes for rare events sampling
• Sampling Rare Conformational Transitions with a Quantum Computer
• EPR-Net: Constructing non-equilibrium potential landscape via a variational force projection formulation
• Generalization and Memorization: The Bias Potential Model
• A Mathematical Framework for Learning Probability Distributions
• Exact Topology of Dynamic Probability Surface of an Activated Process by Persistent Homology
• Reweighted Jarzynski sampling: Acceleration of rare events and free energy calculation with a bias potential learned from nonequilibrium work
• Theory of Moment Propagation for Quantum Dynamics in Single-Particle Description
• Improving the Efficiency of Variationally Enhanced Sampling with Wavelet-Based Bias Potentials
• Variational embedding of protein folding simulations using gaussian mixture variational autoencoders
• Improving the Exploration of High Dimensional Free Energy Landscape by a Combination of Temperature Accelerated Sliced Sampling and Parallel Biasing