In-silico Biopharmaceutical Systems for Provisional Classification of Oral Drugs
Gonzalez-Alvarez M, Pham-The H, Bermejo M, Gonzalez-Alvarez I, Cabrera-Perez MA
Miguel Hernandez University
Classification of drug candidates based on the Biopharmaceutics Classification System (BCS) and the Biopharmaceutics Drug Disposition Classification System (BDDCS) has become an important issue in pharmaceutical researches [1, 2].
Objectives: The main goal of the study was to develop robust in-silico models to classify the solubility, permeability and metabolism properties that define both systems, allowing the definition of a new computational biopharmaceutical filter. The modeling approach for BCS/BDDCS combination is an unexplored area with high relevance for application in both new drug screening and in later drug development stages.
Methods: Three extensive and heterogeneous databases (solubility, permeability and extent of metabolism) and three machine-learning techniques (support vector machine, kappa nearest neighbor and multilayer perceptrons) were used to develop QSPR classification models.
Results: Nine single classification models were selected and three voting systems were constructed. The final consensus models had global accuracies greater than 82% for each property. The in-silico BCS was validated with a dataset of 139 compounds classified by WHO and the in-silico BDDCS was assessed with external dataset of 131 compounds. In the first case, the models correctly classifies 88.4% of class I drugs, 78.3% of class II drugs, 76.6% of class III drugs and 80.8% of class IV drugs. Likewise, the in-silico BDDCS system correctly classifies 78.7% of class I drugs, 80.0% of class II drugs, 87.9% of class III drugs, and 71.4% of class IV drugs. On the basis of both in-silico BCS/BDDCS systems was defined a biopharmaceutical filter that includes eight possible outputs of drugs, drug-like molecules or NMEs.
Conclusions: The results fairly demonstrated the validity of in-silico biopharmaceutical systems for provisional classification of oral drugs in early drug development process.
References:
[1] H. Pham-The, T. Garrigues, M. Bermejo, I. Gonzalez-Alvarez, M.C. Monteagudo, and M.A. Cabrera-Perez. Provisional classification and in silico study of biopharmaceutical system based on caco-2 cell permeability and dose number. Mol Pharm. 10:2445-2461 (2013).
[2] H. Pham-The, I. González-Álvarez, M. Bermejo, T. Garrigues, H. Le-Thi-Thu, and M.A. Cabrera-Pérez. The Use of Rule-Based and QSPR Approaches in ADME Profiling: A Case Study on Caco-2 Permeability. Mol Inf. 32:459-479 (2013).