Molecular enhanced sampling with autoencoders: On-the-fly nonlinear collective variable discovery and accelerated free energy landscape exploration

Macromolecular and biomolecular folding landscapes typically contain high free energy barriers that impede efficient sampling of configurational space by standard molecular dynamics (MD) simulation. Biased sampling approaches improve sampling by driving the simulation along pre-specified collective variables (CVs) to accelerate exploration of configurational space. The success of these methods critically depends on the availability of good CVs of the system. Nonlinear manifold learning techniques can identify such CVs but typically do not furnish an explicit and differentiable relationship between the CVs and atomic coordinates necessary to perform biased MD simulations. In this work, we employ autoencoders to learn nonlinear CVs that are explicit and differentiable functions of atomic coordinates. By interleaving successive rounds of CV discovery and biased sampling, we establish an approach with capacity to simultaneously discover and directly accelerate along data-driven CVs. We demonstrate our approach in simulations of alanine dipeptide and Trp-cage, and have developed an open source framework available at: https://github.com/weiHelloWorld/accelerated_sampling_with_autoencoder.