Yoon, Han Wool2013-11-152013-11-152013-06https://hdl.handle.net/11299/160352University of Minnesota M.S. thesis. June 2013. Major: Biomedical Informatics and Computational Biology. Advisor: Yuk Sham. 1 computer file (PDF); v, 55 pages, appendix A.Accurate prediction of protein pKa`s is important to understand protein electrostatics and functions. Improving the accuracy of pKa prediction using the Poisson-Boltzmann electrostatic model remains an active area of research. The major challenge is to determine the appropriate dielectric constant that best describes the heterogeneous protein environment. The common use of a single large dielectric constant often fails to reproduce large experimental pKa shifts of biological important residues. In this study, I implemented a two steps approach, as described in earlier PDLD/S model, that uses a single low dielectric constant for calculating the intrinsic protein pKa`s when all other ionizable group are neutralized and a single large dielectric constant for evaluating the pKa`s shifts as a result of charge-charge coupling between ionizable groups. This approach is less sensitive to the dielectric constants used and can reliably reproduce the commonly observed protein pKa`s and others with abnormal large pKa shifts.en-USDielectric constantPDLD/SpKa calculation in proteinPoisson Boltzmann Modelprotein pKaThesis or Dissertation