3D-QSAR modeling, molecular docking and ADMET properties of benzothiazole derivatives as α-glucosidase inhibitors

Abstract

Acarbose and voglibose are α -glucosidase inhibitors that have been used for controlling of diabetes mellitus. Unfortunately, these drugs have many side effects. Consequently, the discovery of new agents with high α -glucosidase inhibitory activity and weak side effects becomes of great importance. To that end, a set of 23 benzothiazole compounds have been studied using 3D-QSAR modeling and molecular docking simulation in order to predict new molecules with important α -glucosidase inhibitory activity. CoMFA and CoMSIA models using eighteen compounds in the training set give significant Q2 values of 0.553 and 0.75, and important R2 values of 0.93 and 0.942, respectively. The five remaining molecules wereemployed in order to test the proficiency of 3D-QSAR models and the predicted determination coefficients R2test values are 0.74 and 0.87 for CoMFA and CoMSIA models, respectively. CoMFA and CoMSIA contour maps were afforded much helpful information to determine the preferred and unpreferred regions impacting the activity; as a result, we propose new benzothiazole compounds with important predicted activities. Meanwhile, to guess the interaction between a newly designed compounds and the most active molecule (compound 8) with α-glucosidase receptor (PDB ID: 3L4T), a molecular docking was conducted. The docking outcomes elucidated that the predicted compound X1 was stable in the active pocket of α-glucosidase receptor. Finally, the newly benzothiazole compounds have been assessed for their oral bioavailability and toxicity using ADMET properties and drug likeness. These findings would be of great aid in leading optimization for new drug discovery that can resolve the biggest challenge related of diabetes mellitus.