Reinforced dynamics for enhanced sampling in large atomic and molecular systems
This addresses the challenge of enhanced sampling in molecular dynamics for researchers in computational chemistry and biophysics, offering a novel hybrid method that reduces reliance on precise collective variable selection.
The paper tackles the problem of efficiently exploring configuration space and computing free energy in large atomic and molecular systems by proposing a new approach that combines reinforcement learning with metadynamics, using deep neural networks to train a biasing potential on the fly. It demonstrates the method on systems like alanine dipeptide and polyalanine-10 with 20 collective variables, showing feasibility for handling large sets of variables.
A new approach for efficiently exploring the configuration space and computing the free energy of large atomic and molecular systems is proposed, motivated by an analogy with reinforcement learning. There are two major components in this new approach. Like metadynamics, it allows for an efficient exploration of the configuration space by adding an adaptively computed biasing potential to the original dynamics. Like deep reinforcement learning, this biasing potential is trained on the fly using deep neural networks, with data collected judiciously from the exploration and an uncertainty indicator from the neural network model playing the role of the reward function. Parameterization using neural networks makes it feasible to handle cases with a large set of collective variables. This has the potential advantage that selecting precisely the right set of collective variables has now become less critical for capturing the structural transformations of the system. The method is illustrated by studying the full-atom, explicit solvent models of alanine dipeptide and tripeptide, as well as the system of a polyalanine-10 molecule with 20 collective variables.