Computational methods will be developed and applied to design effective anti-HIV drugs. Monte Carlo (MC) statistical mechanics and Molecular dynamics (MD) simulations will be used to yield quantitative predictions on the structure, energetics, and ligand binding for HIV proteins. The current focus is on understanding and optimizing the activity of non-nucleotide inhibitors of HIV-1 reverse transcriptase (NNRTIs). The MCPRO software system will be used with the OPLS force fields in extensive studies of RT-ligand complexes. Specifically, (1) free-energy perturbations (FEP) and liner response (LRSA) calculations will be carried out for a series of 10-50 analogs of nevirapine and MKC-442 to seek correlations between the computed binding affinities and observed anti-HIV-1 activities. The calculations will be executed in parallel on a multiprocessor computer system. The unprecedentedly large database of results will elucidate the origins of binding variations for RT and also further refine and validate the methodology. (2) Resistance to nevirapine and MKC-442 will be examined through simulations of the effects of key mutations on the binding affinities and structures of the complexes with RT. The appropriate thermodynamic cycle also requires simulations that will provide structures of the unliganded mutant proteins, which have not yet been characterized by diffraction experiments. (3) These results will form a basis for the design of NNRTIs that can defeat the resistance-conferring mutations. Modifications of nevirapine and MKC-442 aimed particularly at the key Y181C mutation will be tested through FEP calculations on the binding of analogs to the native and mutant protein. New NNRTIs will also be developed and optimized through the novel extension of the computational approaches in a combinatorial manner.
Showing the most recent 10 out of 49 publications
