Dual Resolution Membrane Simulations Using Virtual Sites.

Area of Science:

• Computational biology
• Biophysics
• Molecular dynamics

Background:

• All-atomistic (AA) simulations offer high molecular detail but are limited by computational cost.
• Coarse-grain (CG) simulations provide larger time and length scales but sacrifice specific molecular information.
• Existing simulation methods struggle to balance accuracy and efficiency for complex biomolecular systems.

Purpose of the Study:

• To develop a novel hybrid simulation scheme coupling AA and CG resolutions for biomembrane systems.
• To overcome the limitations of purely AA or CG simulations in terms of accessible time/length scales and molecular detail.
• To enable efficient investigation of biomembrane properties at multiple scales.

Main Methods:

• A virtual site (VS) based hybrid scheme was developed to concurrently couple AA and CG resolutions within a single simulation.
• Force fields for AA and CG models were adjusted for compatibility within the hybrid framework.
• The method was applied to simulate lipid bilayer properties and a small lipid vesicle with distinct AA and CG regions.

Main Results:

• The proposed VS hybrid method successfully maintains essential lipid bilayer properties.
• Simulations of a lipid vesicle demonstrated the concurrent use of AA for the inner leaflet/solvent and CG for the outer leaflet/solvent.
• The hybrid approach showed good agreement with expected physical properties.

Conclusions:

• The VS hybrid multiscale method offers a computationally efficient way to simulate biomembranes.
• This approach allows for detailed investigation of membrane properties, especially in systems with large solvent-exposed regions.
• The technique paves the way for studying complex biomolecular processes with improved efficiency and accuracy.