Free energy calculations in molecular dynamics simulations are increasingly applied in drug design to predict relative binding affinities. The electrostatic interactions between the drug molecule, the protein and the solvent, which often govern the binding free energy, are described by atomic charges from standard force field as GAFF, MMFF94, AMBER, OPLS or CHARMM. These atomic charges are obtained with semi-empirical electronic structure methods in vacuum neglecting the polarization of the electron density due the environment or fitted to enthalpy of vaporization data of pure liquids. Here, we propose to apply different partitioning methods (Hirshfeld, MBIS, etc) to calculate the atomic charges of drug molecules directly from the polarized electron density calculated with ab-initio methods in the respective molecular environment. The molecular environment is described implicitly by the SMD solvation moldel or explicitly with the QM/MM methodology averaging the charges over various conformations.<br/>To validate this new method we calculated the hydration free energy of 643 organic molecules in the FreeSolv database through state of the art free energy calculations. We compare free energies calculated with different partitionning scheme in vacuum, using the implicit solvation model or the QM/MM methodology with explicit solvent to previous results using the AM1-BCC atomic charges.

Atom in molecules atomic charges based on different partitioning schemes validated for biomolecular free energy calculations

255th National Meeting (2018)

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.