DESCRIPTION (provided by applicant Adolescent girls in South Africa remain disproportionately affected by the HIV epidemic, despite the availability of several behavioural interventions to reduce the sexual transmission of HIV. Strategies to reduce HIV transmission in this key population would benefit greatly from a better understanding of the sexual networks that drive HIV transmission in adolescent girls. Furthermore, if HIV infection rates in adolescent girls
can be reduced, this could break critical chains of transmission and decrease the spread of HIV in the general population. We hypothesise that complex sexual networks, including mixing between adolescents and community members, drive high HIV incidence in adolescent girls in rural South Africa. To test our hypothesis we will utilize an innovative approach to identify networks of HIV transmission that combine extensive epidemiologic sampling with phylogenetic analyses of HIV-1 sequence data and traditional sexual networking methods. Phylogenetic analysis has recently emerged as one of the most powerful and informative ways to use viral diversity to examine the underlying dynamics of HIV-1 transmission in affected populations. However, these methods have often failed to find linkages in endemic settings; this is likely due to underrepresentation because of sparse sampling of transmission pairs in large populations of HIV infected individuals. One of the major innovations of this proposal is identifying HIV-1 clusters and viral linkages in adolescents and link these to community sequences by extensively sampling HIV infected individuals from a defined geographic area (part of the sub-district of Vulindlela in KwaZulu-Natal), using several epidemiologic approaches to achieve a 'saturating' rate of population coverage. This area is uniquely suited for this research as the study area is rural and geographically well defined; it is one of the highest HIV burden districts in South Afric and CAPRISA has collected detailed data on the structural, behavioural and biological determinants of risks of HIV infection in this community over the last decade. Building on our expertise in HIV surveillance in this area we will i) enhance the representativeness of the phylogenetic analyses by combining novel recruitment strategies including clinic- and school-based recruitment, and respondent-driven sampling for hard-to-reach populations; ii) use high- throughput sequencing, and bioinformatics to identify HIV-1 transmission networks through HIV-1 viral linkages; and iii) include behavioural and biological information and other risk factors. This approach will allow us to characterize HIV transmission clusters and viral linkages in this hyper endemic region, analyze the contribution of HIV clusters to the new infections occurring among adolescent girls, and use this data to improve HIV prevention efforts in this key population for HIV control.