Beyond Ensembles: Simulating All-Atom Protein Dynamics in a Learned Latent Space

Aditya Sengar, Jiying Zhang, Pierre Vandergheynst, PATRICK BARTH

Keywords: representation learning, generative models, latent space dynamics, score-based generative models, dynamical systems, GNNs, autoregressive models

TL;DR: To massively speed up simulations, this AI model compresses a protein's complex all-atom structure into a simple latent space, simulates its movement quickly there, and then reconstructs the full-atom trajectory.

Abstract: Simulating the long-timescale dynamics of biomolecules is a central challenge in computational science. While enhanced sampling methods can accelerate these simulations, they rely on pre-defined collective variables that are often difficult to identify, restricting their ability to model complex switching mechanisms between metastable states. A recent generative model, LD-FPG, demonstrated that this problem could be bypassed by learning to sample the static equilibrium ensemble as all-atom deformations from a reference structure, establishing a powerful method for all-atom ensemble generation. However, while this approach successfully captures a system's probable conformations, it does not model the temporal evolution between them. We introduce the Graph Latent Dynamics Propagator (GLDP), a modular component for simulating dynamics within the learned latent space of LD-FPG. We then compare three classes of propagators: (i) score-guided Langevin dynamics, (ii) Koopman-based linear operators, and (iii) autoregressive neural networks. Within a unified encoder–propagator–decoder framework, we evaluate long-horizon stability, backbone and side-chain ensemble fidelity, and temporal kinetics via TICA. Benchmarks on systems ranging from small peptides to mixed-topology proteins and large GPCRs reveal that autoregressive neural networks deliver the most robust long rollouts and coherent physical timescales; score-guided Langevin best recovers side-chain thermodynamics when the score is well learned; and Koopman provides an interpretable, lightweight baseline that tends to damp fluctuations. These results clarify the trade-offs among propagators and offer practical guidance for latent-space simulators of all-atom protein dynamics.

Primary Area: applications to physical sciences (physics, chemistry, biology, etc.)

Submission Number: 23742