Multiscale models and approximation algorithms for protein electrostatics

TL;DR

This paper discusses boundary-integral methods and multiscale models to improve the accuracy of electrostatic force simulations in biomolecular systems, addressing limitations of traditional continuum models.

Contribution

It introduces new boundary-integral based models, including a multiscale nonlocal continuum theory and nonlinear boundary conditions, enhancing electrostatics modeling for proteins.

Findings

Boundary-integral methods effectively model complex biomolecular electrostatics.

Multiscale nonlocal models capture atomic-scale effects.

Nonlinear boundary conditions improve surface interaction accuracy.

Abstract

Electrostatic forces play many important roles in molecular biology, but are hard to model due to the complicated interactions between biomolecules and the surrounding solvent, a fluid composed of water and dissolved ions. Continuum model have been surprisingly successful for simple biological questions, but fail for important problems such as understanding the effects of protein mutations. In this paper we highlight the advantages of boundary-integral methods for these problems, and our use of boundary integrals to design and test more accurate theories. Examples include a multiscale model based on nonlocal continuum theory, and a nonlinear boundary condition that captures atomic-scale effects at biomolecular surfaces.