A potential anti-Human Immunodeficiency Virus (HIV) agent with novel mode of action is urgently needed to fight against drug resistance HIV. The HIV capsid protein is important for both late and early stages of the viral replication cycle and emerged as a promising target for the developing of small molecule inhibitors of HIV. We design a Pharmacophore and 3D Quantitative structure activity relationship (QSAR) model for HIV Capsid Protein inhibitors, which helps to identify overall aspects of molecular structure that govern activity and for the prediction of novel HIV Capsid inhibitors. The hypothesis was developed with a survival score of 3.6.The features, that is, two aromatic rings, one hydrophobic site and two acceptor regions were present in all the active compounds with good fitness score. Pharmacophore model was then validated by a partial least square and regression-based PHASE 3D QSAR cross-validation. The leave-n-out cross validation for test set (Q2) of the hypothesis is 0.636, the standard deviation (SD) value is 0.338, and the variance ratio (F-test) value is 74.5. Hypothesis also showed a leave-n-out cross validation for training set (R2, 0.928). Interestingly, the predicted activity of true test set compounds was found in the close vicinity of their experimental activity suggesting the methodology used and models generated can be applied to identify potential new chemical entities with better HIV-1 capsid assembly inhibition.Communicated by Ramaswamy H. Sarma.
Keywords: 3D-QSAR; HIV-1 capsid assembly; Pharmacophore mapping; anti-HIV agents.