LAUSR

Abstract Human immunodeficiency virus-1 (HIV-1) glycoprotein 120 (gp120) is one of the key targets for treatment of acquired immunodeficiency syndrome. A large number of inhibitors are being designed for this target in order to find safe and effective drugs. In the present study, quantitative structure activity relationship (QSAR) models established on 128 gp120 indole-based attachment inhibitors have been developed using suitable molecular descriptors. Chemometrics techniques including multiple linear regression (MLR), artificial neural network (ANN) and support vector machine (SVM) methods were used to set up QSAR models in order to explain the structural requirements of HIV-1 gp120 inhibitory activity. The prediction performance of each developed model was evaluated. The results obtained, for both the training and test sets, were encouraging. These results reveal that the predictive power of the SVM model is slightly superior to those of the MLR and ANN models. Further, the docking process was used not only to identify the most probable position and orientation of an inhibitor within the gp120 but also to assess its affinity with this target. This study could help researchers, particularly those working in the field of the pharmaceutical industry, to identify and discover more potent, active, and selective HIV-1 attachment inhibitors. Therefore, the established models could improve, diversify, and accelerate the drug development process and reduce the use of the trial and error approach in the search of new drug targets for the treatment of HIV.