QSAR application for the prediction of compound permeability with in silico descriptors in practical use

Abstract

The parallel artificial membrane permeation assay (PAMPA) was developed as a model for the prediction of transcellular permeation in the process of drug absorption. In our previous report, it was revealed that PAMPA permeability is governed by log P, pK (a), and the hydrogen-bonding ability of compounds. In order to construct a new filtering method for selecting informative compounds from the whole combinatorial library, this study tried to predict PAMPA permeability with in silico descriptors. Log P, pK(a), and polar surface areas (PSA) as a hydrogen-bonding descriptor were calculated by commercially available or free-accessible web programs. Five-fold cross-validations and conventional regression analyses were examined with the training set for the entire 81 combinations with nine log P, three pK(a) and three PSA descriptors. By comparison of statistical indices, four equations were selected and then the model with the best combination of in silico descriptors was determined based on the external validation. The PAMPA prediction equation obtained in this report could be applied for the prediction of both Caco-2 cell permeability and human intestinal absorption of mainly passively-transported drugs.