application) HIV infection in vivo involves many different cell types (uninfected cells permissive for virus replication, infected cells, HIV specific immune cells, and others) that interact with each other and whose numbers, in general, depend on time. The goal of this application is to use mathematical modeling to aid in understanding how such a system can work. The approach is to: i) select mathematical models of HIV/SIV infection based on the criterion that each model's predictions must agree with all significant experimental features of HIV pathogenesis in representative individuals currently available in literature; ii) validate the models selected by finding examples of possible biological mechanisms behind these models; iii) formulate predictions for follow-up experiments to test the models. This """"""""strategy of multiple match"""""""" has been successfully used in the physical sciences for dealing with complex systems of unknown structure, but is novel for HIV research. The investigators will apply this strategy in the two following directions: 1) To elucidate the dominant factors of the virus-T cell interaction in an HIV infection in vivo. 2) To understand the principal factors of HIV evolution in individuals and along the transmission chain.