Coupling Molecular Dynamics and Deep Learning to Mine Protein Conformational Space

Degiacomi, Matteo T.

Abstract

Flexibility is often a key determinant of protein func- tion. To elucidate the link between their molecular structure and role in an organism, computational techniques such as molecular dynamics can be leveraged to characterize their conformational space. Extensive sampling is, however, required to obtain reliable results, useful to rationalize experi- mental data or predict outcomes before experiments are carried out. We demonstrate that a generative neural network trained on protein structures pro- duced by molecular simulation can be used to obtain new, plausible conformations complementing pre- existing ones. To demonstrate this, we show that a trained neural network can be exploited in a pro- tein-protein docking scenario to account for broad hinge motions taking place upon binding. Overall, this work shows that neural networks can be used as an exploratory tool for the study of molecular conformational space.

Citation

Degiacomi, M. T. (2019). Coupling Molecular Dynamics and Deep Learning to Mine Protein Conformational Space. Structure, 27(6), 1034-1040.e3. https://doi.org/10.1016/j.str.2019.03.018

• Journal Article Type: Article
• Acceptance Date: Mar 25, 2019
• Online Publication Date: Apr 25, 2019
• Publication Date: Jun 4, 2019
• Deposit Date: Apr 29, 2019
• Publicly Available Date: Jun 7, 2019
• Journal: Structure
• Print ISSN: 0969-2126
• Electronic ISSN: 1878-4186
• Publisher: Elsevier
• Peer Reviewed: Peer Reviewed
• Volume: 27
• Issue: 6
• Pages: 1034-1040.e3
• DOI: https://doi.org/10.1016/j.str.2019.03.018
• Public URL: https://durham-repository.worktribe.com/output/1303009