HIV infection still has severe health and socio-economic impact and is one of the primary causes of morbidity and mortality all over the world. Antiretroviral therapy has transformed HIV infection from an acute to a chronic disease, increasing life expectancy but also adding to the potential side effects associated with drug therapy and the comorbidity accompanying longevity. Despite the introduction of HAART therapy, the need for new anti-HIV agents is still unabated.<br/>The initiative of the NCI's CADD Group, called the Synthetically Accessible Virtual Inventory (SAVI) aiming at creating up to 1 billion readily synthesized molecules, has opened up new possibilities for discovery of novel anti-HIV agents from the previously unexploited parts of chemical space.<br/>Our comparison of more than 283 million molecules from the SAVI library with about 4,000 pharmaceutical substances from Drug Central has shown no overlap at all. The corresponding overlap analysis of this SAVI database with the most massive freely available library of organic compounds, PubChem, showed about 0.03% overlapping records. These included some compounds with anti-HIV action. These results allow us to consider SAVI as a promising source of novel pharmaceutical agents significantly different from known medicines.<br/>To select new potential anti-HIV agents from SAVI, we applied the computer program PASS (Prediction of Activity Spectra for Substances) and similarity estimates based on MNA (Multilevel Neighborhoods of Atoms) and QNA (Quantitative Neighborhoods of Atoms) descriptors. PASS has been developed and applied for drug design & discovery during the past 25 years, and has been validated in about 500 independent papers reporting on the application of PASS predictions to the discovery of new pharmaceutical agents.<br/>Then, we used molecular modeling techniques to evaluate for dozens of millions of molecules selected by PASS from the SAVI database the possibility of interaction with known HIV-1 targets: reverse transcriptase, protease, and integrase. This combined approach of machine learning, similarity assessment, and molecular modeling led to the selection of dozens of molecules recommended for synthesis and testing for antiretroviral activity.

Analysis of large chemical-biological space aimed at discovery of novel anti-HIV agents

256th National Meeting (2018)

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.