DeepDriveMD: Deep-Learning Driven Adaptive Molecular Simulations for Protein Folding

Abstract

Simulations of biological macromolecules play an important role in understanding the physical basis of a number of complex processes such as protein folding. Even with increasing computational power and evolution of specialized architectures, the ability to simulate protein folding at atomistic scales still remains challenging. This stems from the dual aspects of high dimensionality of protein conformational landscapes, and the inability of atomistic molecular dynamics (MD) simulations to sufficiently sample these landscapes to observe folding events. Machine learning/deep learning (ML/DL) techniques, when combined with atomistic MD simulations offer the opportunity to potentially overcome these limitations by: (1) effectively reducing the dimensionality of MD simulations to automatically build latent representations that correspond to biophysically relevant reaction coordinates (RCs), and (2) driving MD simulations to automatically sample potentially novel conformational states based on these RCs. We examine how coupling DL approaches with MD simulations can fold small proteins effectively on supercomputers. In particular, we study the computational costs and effectiveness of scaling DL-coupled MD workflows by folding two prototypical systems, viz., Fs-peptide and the fast-folding variant of the villin head piece protein. We demonstrate that a DL driven MD workflow is able to effectively learn latent representations and drive adaptive simulations. Compared to traditional MD-based approaches, our approach achieves an effective performance gain in sampling the folded states by at least 2.3x. Our study provides a quantitative basis to understand how DL driven MD simulations, can lead to effective performance gains and reduced times to solution on supercomputing resources.

Transient Unfolding and Long-Range Interactions in Viral BCL2 M11 Enable Binding to the BECN1 BH3 Domain

“…Further, a recent study by Gapsys and de Groot [ 58 ] also highlighted the importance of considering box width and volume for accurate sampling and estimating kinetics and other relevant parameters. We plan to extend our simulations with some of our adaptive sampling techniques as part of a future study given that the conformational changes for the BH3D in the presence of the viral BCL2 takes place in millisecond timescales [ 59 ].…”

AI-Driven Multiscale Simulations Illuminate Mechanisms of SARS-CoV-2 Spike Dynamics

“…Recently, we extended our approach to adaptively run MD simulation ensembles to fold small proteins. This approach, called DeepDriveMD [35], successively learns which parts of the conformational landscape have been sampled sufficiently and initiates simulations from undersampled regions of the conformational landscape (that also constitute "interesting" features from a structural perspective of the protein). While a number of adaptive sampling techniques exist [2,8,30,33,47,67,68], including based on reinforcement learning methods [39], these techniques have been demonstrated on prototypical systems.…”

Including residual contact information into replica-exchange MD simulations significantly enriches native-like conformations

“…It has a sequence length of 35 residues and forms a three-helix structure. Similar to Trp-Cage, VHP can also achieve folding times in the order of μs [53,54] and has a folding temperature of approximately 339 to 342 K [55].…”