This project aims to parameterize an energy model with the goal of developing a fast method for predicting binding affinities of HIVP inhibitors. This method will be used for in silico compound screening to discover new potential anti-HIV drug candidates. The project also aims to develope a method of optimizing the charges of local parts of a ligand while keeping the rest of the charges roughly constant, rather than attempting to modify all of the ligand's charges towards an optimum, as done in previous approaches. The method developed here will also be computationally faster than existing approaches.