An open dataset of Plasmodium vivax genome variation in 1,895 worldwide samples

Abstract

This report describes the MalariaGEN Pv4 dataset, a new release of curated genome variation data on 1,895 samples of Plasmodium vivax collected at 88 worldwide locations between 2001 and 2017. It includes 1,370 new samples contributed by MalariaGEN and VivaxGEN partner studies in addition to previously published samples from these and other sources. We provide genotype calls at over 4.5 million variable positions including over 3 million single nucleotide polymorphisms (SNPs), as well as short indels and tandem duplications. This enlarged dataset highlights major compartments of parasite population structure, with clear differentiation between Africa, Latin America, Oceania, Western Asia and different parts of Southeast Asia. Each sample has been classified for drug resistance to sulfadoxine, pyrimethamine and mefloquine based on known markers at the dhfr, dhps and mdr1 loci. The prevalence of all of these resistance markers was much higher in Southeast Asia and Oceania than elsewhere. This open resource of analysis-ready genome variation data from the MalariaGEN and VivaxGEN networks is driven by our collective goal to advance research into the complex biology of P. vivax and to accelerate genomic surveillance for malaria control and elimination.

Keywords: data resource; genomic epidemiology; genomics; malaria; plasmodium vivax.

Figure 1.. Population structure.

( A) First two components of a genome-wide principal coordinate analysis. Each point represents one of 1,072 QC pass samples coloured according to country groupings ( Table 1): Latin America (green, n=159); Africa (red, n=137); Western Asia (orange, n=47); West south-east Asia (blue; n=127); East south-east Asia (purple; n=277); Maritime south-east Asia (pink; n=76); Oceania (brown; n=208); Unassigned samples (grey; n=41). This shows the genetic separation of samples into seven distinct geographic clusters. This also shows that samples that have not been assigned to a region look distinct from those from the seven regions. After removal of the 41 unassigned samples we have an analysis set of 1,031 samples. ( B) Genome-wide unrooted neighbour-joining tree showing population structure across all sites from the seven regions (1,031 analysis set samples), with sample branches coloured as in A. This shows that maritime Southeast Asia has large numbers of very highly related parasites and clear relatedness between samples is also present in some samples from Latin America and Africa.