Abstract

We systematically investigate the interactions for HIV drug resistance and cross-resistance of three protease inhibitors. Using Bayesian probabilistic models, we first detect mutation combinations related to drug resistance in the HIV-1 protease or reverse transcriptase (RT) of patients with single, double and triple drug treatments respectively, and then infer detailed interaction structures of these mutations. Moreover, we compare the resistance patterns of single-drug treatments and multiple-drug treatments and study the selection patterns of mutations from single-drug treatments and multiple-drug treatments, using molecular dynamics simulation and free energy calculations, in order to draw inference for giving advice on the design of multiple-drug treatments.