Membrane Potential: Accuracy and Reproducibility of Molecular Dynamics Simulations

Area of Science:

• Membrane biophysics
• Computational chemistry
• Molecular dynamics simulations

Background:

• The membrane dipole potential (Ψd) is crucial for ion transport and protein function.
• Accurate prediction of Ψd modifications is essential for drug design.
• Molecular dynamics (MD) simulations are a valuable tool but sensitive to setup and parameters.

Purpose of the Study:

• To systematically evaluate how simulation setup and parameterization influence calculated membrane dipole potential.
• To compare different force fields, water models, and small-molecule parameterization protocols.
• To identify reliable MD protocols for predicting Ψd modifications.

Main Methods:

• Utilized CHARMM36m and AMBER (Lipid21) force fields.
• Investigated effects of system composition, box size, and water models.
• Compared CGenFF, ffTK-refinement, GAFF2, and Espaloma for flavonoid parameterization.

Main Results:

• CHARMM force field showed higher sensitivity to box composition and finite-size effects than AMBER.
• 3-site water models are reliable for predicting relative potential changes.
• Espaloma and ffTK-refinement significantly improved accuracy for flavonoid parameterization compared to CGenFF.

Conclusions:

• Provides practical recommendations for establishing reliable MD protocols to predict Ψd modifications.
• Highlights Espaloma as a promising automated approach for accurate Ψd prediction.
• Emphasizes the importance of force field choice and parameterization strategy in MD simulations.