ANN modeling of the penetration across a polydimethylsiloxane membrane from theoretically derived molecular descriptors

A quantitative structure-permeability relationship was developed using Arti ficial Neural Network (ANN) modeling to study penetration across a polydime thylsiloxane membrane. A set of 254 compounds and their experimentally deri ved maximum steady state flux values used in this study was gathered from t he literature. A total of 42 molecular descriptors were calculated for each compound. A genetic algorithm was used to select important molecular descr iptors and supervised ANN was used to correlate selected descriptors with t he experimentally derived maximum steady-state flux through the polydimethy lsiloxane membrane (log J). Calculated molecular descriptors were used as t he ANN's inputs and log J as the output. Developed model indicates that mol ecular shape and size., inter-molecular interactions, hydrogen-bonding capa city of drugs, and conformational stability could be used to predict drug a bsorption through skin. A 12-descriptor nonlinear computational neural netw ork model has been developed for the estimation of log J values for a data set of 254 drugs. Described model does not require experimental parameters and could potentially provide useful prediction of membrane penetration of new drugs and reduce the need for actual compound synthesis and flux measur ements. (C) 2001 Elsevier Science B.V. All rights reserved.