Continuum dielectric modelling of electrostatics interactions in macro
molecules provides a valuable tool in the study of structure-function
relationships, but falls short of providing consistently accurate calc
ulated pK(a)s. It is suggested that the model can be significantly imp
roved with the inclusion of a term that estimates the entropy associat
ed with first hydration shell solvent ordering, with reference to comp
uted results for cysteines in DsbA and thioredoxin, and aspartic and g
lutamic acids in a number of proteins, The modification is based on th
e geometry of charge burial and an hydration number, which is adjustab
le (by fit to experiment), and is uniform within each class of ionizab
le group studied, The potential for further development is clear withi
n this framework, since experiment and simulation can furnish non-adju
stable, ionizable group-specific, hydration numbers.