Treffer: OpenCafeMol: A coarse-grained biomolecular simulator on GPU with its application to vesicle fusion.

Title:
OpenCafeMol: A coarse-grained biomolecular simulator on GPU with its application to vesicle fusion.
Authors:
Murata Y; Department of Biophysics, Graduate School of Science, Kyoto University, Kyoto, Japan., Niina T; Laboratory for Computational Molecular Design, RIKEN Center for Biosystems Dynamics Research (BDR), Kobe, Hyogo, Japan., Takada S; Department of Biophysics, Graduate School of Science, Kyoto University, Kyoto, Japan. Electronic address: takada@biophys.kyoto-u.ac.jp.
Source:
Biophysical journal [Biophys J] 2026 Jan 20; Vol. 125 (2), pp. 420-431. Date of Electronic Publication: 2025 Jul 12.
Publication Type:
Journal Article
Language:
English
Journal Info:
Publisher: Cell Press Country of Publication: United States NLM ID: 0370626 Publication Model: Print-Electronic Cited Medium: Internet ISSN: 1542-0086 (Electronic) Linking ISSN: 00063495 NLM ISO Abbreviation: Biophys J Subsets: MEDLINE
Imprint Name(s):
Publication: Cambridge, MA : Cell Press
Original Publication: New York, Published by Rockefeller University Press [etc.] for the Biophysical Society.
MeSH Terms:
Molecular Dynamics Simulation* , Membrane Fusion* , Software* , Computer Graphics*, SNARE Proteins/chemistry ; SNARE Proteins/metabolism ; Protein Folding
Substance Nomenclature:
0 (SNARE Proteins)
Entry Date(s):
Date Created: 20250714 Date Completed: 20260121 Latest Revision: 20260122
Update Code:
20260123
DOI:
10.1016/j.bpj.2025.07.012
PMID:
40653709
Database:
MEDLINE

Weitere Informationen

There has been an increasing demand for longer-timescale molecular dynamics (MD) simulations of larger biomolecular systems. To meet these demands, using the C++ API of OpenMM, we developed a fast and flexible MD software, OpenCafeMol, for residue-resolution protein and lipid models that shows high performance on graphics processing unit (GPU) machines. We validated OpenCafeMol for folding small proteins, lipid membrane dynamics, and membrane protein structures. Benchmark tests of the computation times showed that OpenCafeMol with one GPU for proteins and lipid membranes is approximately 100 and 240 times faster than the corresponding simulations on a typical CPU machine (eight cores), respectively. Taking advantage of the high speed of OpenCafeMol, we applied it to two sets of vesicle fusion simulations; one driven by force and the other coupled with conformational dynamics of a soluble N-ethylmaleimide-sensitive factor attachment protein receptor (SNARE) complex. In the latter MD simulation at a high temperature resulted in vesicle docking and pore formation, followed by fusion, which are coupled with local folding of linkers in the SNARE complex. This opens up a new avenue to study membrane-fusion mechanisms via MD simulations. The source code for OpenCafeMol is fully available.
(Copyright © 2025 Biophysical Society. Published by Elsevier Inc. All rights reserved.)

Declaration of interests The authors declare no competing interests.