We propose to determine atomic partial charges for commonly occurring amino acids, the building blocks of proteins, from experimental data. These charges are necessary parameters for the force fields used in molecular modeling studies of proteins. Previously, such charges have been derived using quantum mechanical and other theoretical techniques. Since the energies, and thus conformations, of proteins are believed to be heavily influenced by charge-charge interactions within these molecules, it is very important that the charge parameters used in molecular mechanics and dynamics calculations be as accurate as possible. Using recently derived methods, we are now able to derive the necessary charges from high accuracy x-ray crystallographic data. The charges so determined will serve a number of purposes, including: 1) They will be used in an improved force field for future molecular modeling studies of proteins; and 2) They will serve as a control set of data with which to compare the results of theoretical charge calculations. By comparing charge sets, we can improve the theoretical techniques (determination of an optimal basis set, accounting for the effects of hydrogen bonding, etc.), ultimately making these methods more reliable for charge determinations in those cases for which an experimental determination is infeasible.
