Prediction of water–octanol partition coefficients using theoretical descriptors derived from the molecular surface area and the electrostatic potential

Abstract

Octanol–water partition coefficients are frequently used in quantitative structure–activity relationships. A correlation based on computed theoretical descriptors is presented for the prediction of octanol–water partition coefficients (Pow). An ab initio SCF approach was used to compute the molecular descriptors at the HF/6-31G* level. It was shown that only three theoretical parameters representing a cavity term, a dipolarity/polarisability term and a hydrogen bonding term were needed for the correlation. The corresponding parameters were deduced from the molecular surface area, the surface electrostatic potential and the spatial minima of the electrostatic potential, respectively. The predictive power of log Pow was demonstrated on a number of molecules which have biological activity.