Simulation_of_P__falciparum_Complexity_of_Infection__COI_

Many P. falciparum infections, particularly in Africa, are a mixture of more than one parasite strain due to repeated mosquito inoculations, which can themselves host complexity. MalariaGEN has several approaches for accommodating within-sample heterozygosity in analyses, all of which assume that the strains comprising the infection yield sequencing reads in proportion to their respective abundance in the sample, and that this relationship holds across mixture ratios. For example, an important metric related to the inbreeding coefficient, Fws, relies on within-sample read proportions to estimate heterozygosity. We have a poor understanding of how well read ratios estimate DNA ratios, and how this might depend on the degree of mixing. Within-sample heterozygosity also leads to inaccurate de novo assemblies, and thus our current pipeline excludes samples with polymorphism in areas being assembled. This results in a substantial loss of data. It would therefore be valuable to determine the exact thresholds of mixture at which assembly remains accurate, and potentially salvage currently discarded samples. For information on this project, contributing investigators and studies, sample details, and data sharing policies please visit http://www.malariagen.net/node/287