Implicit solvents for the polarizable atomic multipole AMOEBA force field: application to small molecules, proteins, and nucleic acids

Computationally modeling molecules and molecular interactions can provide a broadly useful supplement to wet bench experiments in illustrating chemical structure and function. Molecular dynamics simulations are designed to show how molecules move and interact in a given environment. Accurate representation of the environment is essential for meaningful simulations and, for biomolecules, this includes representing bulk water. The gold standard for simulation in water is to explicitly include the hundreds or even thousands of individual water molecules surrounding the molecule of interest. Keeping track of the movement and positioning of so many molecules can quickly become costly and render simulations inefficient or even intractable. Implicit models of solvent that represent water as a continuum provide an attractive alternative with the potential to dramatically improve simulation speed and scale.

Here the development and parameterization of a family of implicit solvent models for use with the atomic multipole optimized energetics for biomolecular applications (AMOEBA) polarizable force field are described. First, three implicit solvent models for use with small molecules are implemented. Next, one of the models is expanded for use with biomolecules, specifically proteins and nucleic acids. The corrections needed to account for volume scale differences between small molecules and biomolecules are described. Finally, molecular dynamics simulations are conducted for a diverse set of proteins and nucleic acids to better understand the current state of the biomolecular implicit solvent model. This initial development of AMOEBA implicit solvent models lays the groundwork for simulation of arbitrary biomolecules on biologically relevant timescales.