A general formalism for a computational stimulation of the biomolecular functions at a microscopic level has been developed. This method was achieved by a combination of the boundary element method and PDLD model in the case of nonzero ionic strength. Calculations with our method have been performed both on a simple model and on realistic problems. The latter concerns the solvation energies of foul residues in bovine pancreatic trypsin inhibitor (BPTI) and the pK-shift of His-64 of mutant subtilisin BPN’. We focus our analysis on evaluating the influence of the parameter such as polarizability, charge, ionic strength, solvent screening, etc.,on the total solvation energies. Our calculation confirms that bulk solvent screening effects on protein induced dipoles can be simply ignored, introducing a relative error less than 5%, while the polarization of protein induced dipoles on the solvent accounts for up to -10 kcal/mol and must be considered, and that the change of solvation energies is not sensitive to the parameters of charge, polarizability, and ionic strength. Moreover, it is found that the effects of protein induced dipoles that are more than 9 Angstrom away from the sites of interest can be ignored and that all atom partial charges should be considered.