Modeling the levels, trends, and connectivity of malaria transmission using genomic data from a health facility in Thies, Senegal

Molecular data and analysis outputs are being integrated into malaria surveillance efforts to provide valuable programmatic insights for national malaria control programs (NMCPs). A plethora of studies from diverse geographies have demonstrated that malaria parasite genetic data can be an important tool for drug resistance monitoring, species identification, outbreak analysis, and transmission characterization. Despite many successful research efforts, there are still important knowledge gaps hindering practical translation of each of these use cases for NMCPs. Here, we leverage epidemiological modeling and time-series data of 2035 genetic sequences collected in Thies, Senegal from 2006-2018 to provide a quantitative and setting-specific assessment of the levels, trends, and connectivity of malaria transmission. We also identify the genetic features that are the most informative for inferring transmission in Thies, such as the fraction of the population with multiple infections and the persistence of parasite lineages across multiple transmission seasons. The model fitting and uncertainty quantification framework also reveals a significant decrease in the level of malaria transmission around 2013. This difference coincides with a large-scale drought and bed net campaign by the NMCP and USAID and is independently corroborated by geo-spatial models of incidence in Thies. We find that genetically identical samples are more likely to be geographically clustered even at the neighborhood scale; and moreover, these lineages propagate non-randomly around the city. Our approach and results provide quantitative guidance for the interpretation of malaria parasite genetic data from Thies, Senegal and indicates the value of increased malaria genomic surveillance for NMCPs.