Assessing molecular docking tools for relative biological activity prediction: a case study of triazole HIV-1 NNRTIs

Molecular docking is a technique widely used in drug design. Many studies exist regarding the general accuracy of various docking programs, but case studies for a given group of related compounds are rare. In order to facilitate identification of novel triazole HIV-1 non-nucleoside reverse transcriptase inhibitors (NNRTIs), several docking and scoring programs were evaluated for their ability to predict relative biological activity of 111 known 1,2,4-triazole and 76 other azole type NNRTIs. Glide, FlexX, Molegro Virtual Docker, AutoDock Vina, and Hyde were used. Different protocols, settings, scoring functions, and interaction terms were analyzed. We have found that the programs performance was dependent on the data set, indicating the importance of choosing good quality target data for any comparative study. The results suggest that after optimization and proper validation, some of the molecular docking programs can help in predicting relative biological activity of azole NNRTIs.