Effective Approximation of Molecular Volume Using Atom-Centered Dielectric Functions in Generalized Born Models

Area of Science:

• Computational chemistry
• Biophysics
• Molecular modeling

Background:

• Generalized Born (GB) theory offers efficient implicit solvent treatment for protein simulations.
• Accurate description of the dielectric boundary is crucial for GB models.
• Existing GB models often use computationally simple van der Waals (vdW) surfaces instead of more physical molecular surfaces (MS).

Purpose of the Study:

• To develop and validate a new Generalized Born with Surface area (GBSW) model, GBSW/MS2, that approximates molecular surfaces efficiently.
• To assess the accuracy of GBSW/MS2 in reproducing Born radii and desolvation effects.
• To optimize GBSW/MS2 parameters for improved simulation of protein conformational equilibria.

Main Methods:

• Developed GBSW/MS2 using atom-centered dielectric functions to approximate molecular volume.
• Compared GBSW/MS2 Born radii against Poisson-Boltzmann calculations.
• Analyzed potentials of mean force for hydrogen-bonding and charge-charge interactions.
• Optimized GBSW/MS2 parameters using solvation free energies and conformational data of model peptides.

Main Results:

• GBSW/MS2 achieves efficiency comparable to vdW-based GBSW models.
• GBSW/MS2 shows high correlation (0.95) with Poisson-Boltzmann-derived Born radii.
• GBSW/MS2 correctly captures key desolvation peaks, indicating accurate MS representation.
• The optimized GBSW/MS2 force field reasonably reproduces structures and stabilities of model peptides.

Conclusions:

• GBSW/MS2 provides an efficient and accurate method for implicit solvent treatment using molecular surfaces.
• The model shows promise for reliable simulations of protein conformational equilibria.
• Further developments and addressing limitations will enhance its applicability in biophysical studies.