(409h) Deep Learning of Low-Dimensional Latent Space Molecular Simulators

The long-time microscopic evolution of molecular systems is governed by the leading eigenfunctions of the transfer operator that propagates the system dynamics through time. The low-dimensional latent space defined by these eigenfunctions parameterize the slow manifold to which the system dynamics are constrained to evolve. A set of three deep neural networks of different architectures trained over short molecular simulation trajectories provides a means to (i) learn the leading transfer operator eigenfunctions, (ii) propagate the dynamics within the encoded latent space, and (iii) decode the latent space back to the all-atom coordinate space. This technique offers a means to train numerical simulators to conduct molecular simulations and estimate thermodynamic and kinetic observables at orders-of-magnitude lower cost than conventional molecular dynamics calculations.