Prediction of Distribution Coefficient from Structure. 2. Validation of Prolog D, an Expert system

Abstract

Prolog D is a program that formalizes, in a controllable and reproducible manner, an algorithm developed to predict distribution coefficients of ionizable compounds at a given pH and varying counterion concentrations. Its predictive power has been evaluated with experimental log D values measured under standard conditions of buffers and ionic strength. Calculations were performed with the three different options for the estimation of partition coefficients (log P) implemented in the program. Considering the diversity of test compounds as well as the present state of the art in log P and pKa predictions, Prolog D proved to be very efficient and can be used as a tool to provide lipophilicity data. Prediction patterns and correlations with the observed data are of almost equal quality for all options, permitting acceptable results for 80% of the data.