Towards computational prediction of Biopharmaceutics Classification System: a QSPR approach

Today classification of drug candidates on the Biopharmaceutics Classification System (BCS) has become an important issue in pharmaceutical researches. In this work, we provide a potential in silico approach to predict this system using two separately classification models of Dose number and Caco-2 cell permeability. 18 statistical linear and nonlinear models have been constructed based on 803 0-2D Dragon and 126 Volsurf+ molecular descriptors to classify the solubility and permeability properties. The voting consensus model of solubility (VoteS) showed a high accuracy of 88.7% in training and 92.3% in test set. Likewise, for the permeability model (VoteP), accuracy was 85.3% in training and 96.9% in test set. A combination of VoteS and VoteP appropriately predicts the BCS class of drugs (overall 73% with class I precision of 77.2%). This consensus system predicts the BCS allocations of 57 drugs appeared in the WHO Model List of Essential Medicines with 87.5% of accuracy. A simulation of a biopharmaceutical screening assay has been proved in a large data set of 37,377 compounds in different drug development phases (1, 2, 3 and launched), and NMEs. Distributions of BCS forecasts illustrate the current status in drug discovery and development. It is anticipated that developed QSPR models could offer the best estimation of BCS for NMEs in early stages of drug discovery.