Assessing ligand efficiencies using template-based molecular docking and Tabu-clustering on tetrahydroimidazo-[4,5,1-jk][1,4]-benzodiazepin-2(1H)-one and-thione (TIBO) derivatives as HIV-1RT inhibitors

A template-based flexible docking simulation followed by ‘Tabu-clustering’ was performed on a series of 38 TIBO derivatives as HIV-1 reverse transcriptase (HIV-1 RT) inhibitors. Four different templates of the Cl-TIBO (1-REV) were created and used as reference templates for docking and aligning. On the basis of the optimal conformation of the ligands, when fitting to the template, the respective scoring functions were obtained; different ligand efficiencies were evaluated and analysed. Statistical modelling using artificial neural network (ANN: r 2 = 0.922) and multiple linear regression method (MLR: r 2 = 0.851) showed good correlation between the biological activity, binding affinity, and different ligand efficiencies of the compounds, which suggest the robustness of the template-based binding conformations of these inhibitors. Our studies suggest that, template-based docking followed by ‘Tabuclustering’ will give a better alignment of inhibitors with respect to the crystal coordinates and enhance the docking efficiency. Also, our study indicates that scoring functions based on 3D symmetry analysis along with heavy atoms count serve as a valuable tool for estimating the efficiency of the ligands. Thus, this is a novel method based on heavy atoms count predicting the binding affinity of the TIBO group of inhibitors, so that their therapeutic utility can be enhanced.