Development of 2D-QSAR Models and Molecular Docking of Protein Tyrosine Phosphatase 1B Inhibitors

Abstract

Quantitative structure-activity relationships (QSARs) are widely used in drug discovery and design to quantitatively analyze the relationships between the structures and biological activities of compounds. The present study aimed to develop a two-dimensional (2D) QSAR and molecular docking studies to predict the biological activities of protein tyrosine phosphatase 1B (PTP1B) inhibitors. A QSAR was carried out to study a series of fifty-three (53) compounds based on protein tyrosine phosphatase 1B (PTP1B) inhibitors. The study was performed using principal components analysis (PCA), multiple linear regression (MLR) and multiple non-linear regression (MNLR) to predict unambiguous QSAR models of studied compounds toward PTP1B inhibitory activity. Molecular docking was used to elucidate the inhibitory mechanisms of the most active compound from the data set against PTP1B. The statistical results of the MLR and MNLR indicate that the determination coefficients (R²) were similar (R² = 0.796). To validate the predictive power of the resulting models, the determination coefficients of external validation were 0.815 and 0.639 for the MLR and MNLR, respectively. These results showed that both models possess favorable estimation stability and good prediction power. The docking of the most active compound 46 showed many hydrogen bond formations with the active site residues ASP (A:48) and TYR (A:46) of PTP1B. The study successfully developed a simple, convenient quantitative structure-activity relationship (QSAR) model that can be used to screen chemical databases or design new PTP1B inhibitor-derived compounds.