Brief Introduction

Despite considerable progress, treatment of AIDS still faces multiple challenges. To date, no truly effective drug able to eliminate HIV has been developed. Furthermore, HIV is highly variable and can quickly acquire resistance against any drug with which it is confronted. Therefore, there is a constant demand to develop new, effective and affordable anti-HIV drugs. In the past decades, tens of millions of chemical compounds have been deposited in public databases. Screening these huge databases for new anti-HIV drugs through experimental methods is a tedious, expensive and time-consuming process. The time and money-saving way is that all compounds in the database are firstly filtered by the computational analysis of the anti-HIV potential, then evaluated by experiment. Therefore, a rapid and accurate computational method is urgently required for predicting anti-HIV activity of chemical compounds.

In this project, we developed a web server called Anti-HIV Predictor for predicting anti-HIV activity of given compounds with high performance (Accuracy > 96% and AUC > 0.96), which enables us to screen tens of millions of compounds and discover new anti-HIV agents. The web server was applied to drug repositioning for anti-HIV therapy by screening 1836 approved drugs.