Chemometric classification of HIV-1 protease inhibitors

This work presents a chemometric classification for a set of 24 known HIV-1 protease inhibitors, based on two pattern recognition methods widely used in quantitative structure-activity (QSAR) studies: soft independent modeling of class analogy (SIMCA) and Kth nearest neighbors (KNN) with the aim of revealing the most relevant structural and physicochemical variables for the design of novel HIV-1 protease inhibitors. The chosen HIV-1 protease inhibitors have inhibitory activity (K-i) ranging from 0.03 to 19,000 nM. Useful information on quantitative structure-activity relationship for this set of inhibitors was found, our model is capable to classify a test set of 5 inhibitors with 80% of correction, and also this methodology had identified significant chemical variables related to anti-HIV-1 protease activity. Inhibitors with high chemical similarity tend to masquerade the chemometric classification (C) 2000 John Wiley & Sons, Inc.