QSAR Modeling of Styrylquinoline Derivatives as HIV-1 Integrase Inhibitors

Background: AIDS is a complicated disease, and the underlying complication makes a total cure impossible. This demands the vigorous need for suitable anti-HIV agents. Styrylquinoline, a quinolone derivative, emerged as a potent HIV-IN inhibitor. Objective: This study aims to construct an easily transferable and reproducible model that relates the biological activities of styrylquinoline derivatives to their molecular descriptors. Methods: 2D Quantitative structure-activity relationship (QSAR) studies were carried out on a series of 36 styrylquinoline derivatives. Results: The technique of recursive feature elimination with random forests was used to select the descriptors rich in information regarding biological activity. The selected descriptors were used in QSAR studies based on multiple linear regression (MLR) and multiple nonlinear regression (MNLR). The performance of models was evaluated by internal and external validations. The values of R²pred and Q²LOO for the MLR model are 0.814 and 0.713, respectively, while the MNLR model has R²pred and Q²LOO values of 0.810 and 0.699, respectively. Conclusion: The information obtained from 2D-QSAR models will aid in gaining a better understanding of the structural requirements for creating novel HIV-IN inhibitors.