A central goal of the Protein Structure Initiative: Biology (PSI:Biology) is to determine structures for all proteins encoded by the human exome using a combination of experimental and computational methods. From the time the PSI first began until very recently, the prior chemical information in protein force fields used during X-ray crystallography refinement and homology modeling were limited by two simplifying approximations. First, atomic charge density was defined to be spherically symmetric, which neglects aspherical atomic features such as oxygen and nitrogen lone-pair positions that drive hydrogen bonding in protein a-helix and �-sheet secondary structures. Second, electron density was defined to be fixed with respect to the electrostatic characteristics of the environment. This neglects, for example, the electrostatic response of amino acids as they move between the polar solvent environment and fold into the hydrophobic protein core. Over the last decade, a state-of-the-art polarizable force field called AMOEBA has emerged that eliminates both approximations. In combination with particle-mesh Ewald (PME) summation, this opens the door to new computational methods that have the potential to improve most, if not all, PSI:Biology protein structures from both X-ray crystallography experiments and homology modeling. Here we present results that establish this approach routinely improves protein structures refined against low-resolution diffraction data (4 �) to a level of quality consistent with mid to high-resolution diffraction (2 �), which is essential for structure-based drug discovery.

Protein structure refinement with a polarizable force field and PME

246th National Meeting (2013)

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