Recurrent malaria infections in endemic regions do not generate sterilizing immunity, suggesting extensive antigenic diversity. The var multigene family provides the major antigenic variation at the blood stage in Plasmodium falciparum malaria. With 50-60 var genes per genome and prevalent meiotic recombination, it is unclear whether immune selection can act as a dominant force in structuring var repertoires of local populations. Using simulated data and empirical data from Ghana, we investigate the role of immune selection in shaping the population structure of var repertoires and repertoires over time.
We extend the existing theory on immune selection and develop an agent-based immune selection model to encompass the vast diversity of var genes in endemic regions. We then develop two neutral models that retain the demography of the transmission system but eliminate specific competition. We propose single and multi-layer network analyses of genetic similarity to identify signatures of dominant processes underlying strain structure.
Static and dynamic networks of genetic similarity both reveal distinctive properties of immune selection that depart from neutral expectations. At the repertoire level, an ensemble of network properties is required to distinguish immune selection from neutrality. When inspected in time, modules of repertoires under immune selection persist much longer than those of neutrality. The population structure of the empirical var tag data from Ghana resembles theoretical expectations under immune selection, with long temporal persistence of var repertoire modules.
These findings support a significant dynamical role of competition for hosts in the generation and maintenance of P. falciparum strains in high transmission regions. The resulting coexistence and diversity of repertoires would enable the large reservoir of asymptomatic infections. As such, monitoring var gene diversity and structure in responses to control efforts becomes central to understanding malaria epidemiology.