Challenges of conventional iterative all-atom and coarse-grained multiscale molecular dynamics

Area of Science:

• Biomolecular simulations
• Computational biophysics
• Molecular dynamics

Background:

• Multiscale molecular dynamics (MD) simulations offer a way to study biomolecular systems at different resolutions.
• Iterating between all-atom (AA) and coarse-grained (CG) representations is a common approach in multiscale simulations.

Purpose of the Study:

• To evaluate the biomolecular dynamics behaviors of conventionally iterating between AA and CG MD simulations.
• To identify practical applications for iterative AA and CG simulations without constraints or model modifications.
• To assess the limitations of this iterative multiscale MD (iMMD) workflow.

Main Methods:

• Implementation of an iterative multiscale MD (iMMD) simulation workflow in OpenMM.
• Evaluation of the iMMD workflow on four representative systems: soluble protein folding and membrane protein interactions (protein-protein and protein-lipid).
• Utilizing readily available AA and CG force fields.

Main Results:

• Iterative AA and CG simulations can assist soluble proteins in exiting metastable states and achieving folding, potentially due to random structural distortions during cycling.
• The iMMD workflow shows greater efficacy in accelerating complex lipid mixing for membrane-bound protein systems compared to sampling protein conformational space.
• The study explores practical usages and limitations of iterative AA and CG simulations.

Conclusions:

• Iterative multiscale simulations are more reliably applied to accelerating lipid mixing in membrane systems than for sampling protein conformational space.
• The developed iMMD workflow provides insights into the practical applications and limitations of conventional AA-CG simulation iteration.
• The iMMD workflow is publicly available for further research.