Are Crystallinity Parameters Critical for Drug Solubility Prediction?

Abstract

Solubility is one the most important physicochemical properties in various stages of drug discovery and development. Different quantitative structure–property relationship models, using diverse parameters, have been proposed to predict the aqueous and non-aqueous solubilities of pharmaceutical compounds. In this study, the effects of these parameters, especially crystallinity parameters, were examined for prediction of the solubility of drug and drug-like molecules in water and octanol. The descriptors, i.e. octanol–water partition coefficient (log10 P), topological polar surface area, Abraham solvation parameters and crystallinity parameters, were computed or extracted from the literature. The results demonstrate that log10 P is the key parameter in aqueous solubility prediction and none of the other descriptors was superior to the others. However, the models, which are composed of calculated parameters such as the Abraham solvation parameters, can be applied to virtual compounds in drug discovery. On the other hand, crystallinity parameters are crucial in predicting the solubility of pharmaceutical compounds in octanol, and these parameters should be considered in developing models.