A strategy for reducing gross errors in the generalized Born models of implicit solvation

Area of Science:

• Computational Chemistry
• Biophysics
• Molecular Modeling

Background:

• The canonical generalized Born (GB) formula estimates biomolecular electrostatic solvation energies (ΔG(el)) using effective Born radii.
• Accurate ΔG(el) is crucial for understanding biomolecular behavior.

Purpose of the Study:

• To identify and rectify the sources of significant errors in the canonical GB formula, even with accurate effective Born radii.
• To develop an improved analytical Green function for more accurate electrostatic solvation energy calculations.

Main Methods:

• Analysis of exact analytical solutions to the Poisson equation (PE) for idealized nonspherical geometries.
• Identification of two distinct spatial modes in PE solutions, one of which is missed by the canonical GB Green function.
• Development of a new analytical Green function incorporating effective Born radii and their gradients.

Main Results:

• The canonical GB formula exhibits gross errors in pairwise atomic charge interactions due to missing a spatial mode in its Green function.
• The proposed functional form captures both spatial modes of the PE solution, improving accuracy for nonspherical shapes.
• Tests on biomolecular structures show the new functional form reduces gross pairwise errors by up to an order of magnitude for small molecules and by a factor of 2 for larger proteins.

Conclusions:

• The proposed analytical Green function offers a significant improvement over the canonical GB formula for electrostatic solvation energy calculations.
• This new method enhances the representation of nonspherical molecular shapes, leading to more reliable predictions.
• The findings have implications for accurate molecular modeling and simulation of biomolecules.