View Program - 2022 Population, Evolutionary, and Quantitative Genetics Conference

Intensive malaria elimination efforts to counter the spread of multidrug-resistant Plasmodium falciparum (P.f.) have led to a dramatic decrease in malaria cases in the Greater Mekong Subregion over the past decade, with a corresponding reduced parasite effective population size (Ne) and increased population structure. Estimates of parasite population demography are critical for monitoring sources and sinks of malaria transmission and the impact of targeted interventions in reducing malaria. Identity-by-descent (IBD) has been widely used in human studies to estimate Ne and population structure in recent time scales. However, the strong directional selection related to antimalarial drug resistance in P.f. may violate the assumptions made in human-oriented IBD-based tools, potentially leading to inaccurate estimation of parasite population demography. Here, we combined the power of coalescent (msprime) and forward simulators (SLiM) to simulate the genealogical trees with varying directional selection parameters, implemented a tree-based true IBD-estimation algorithm (which circumvents the bias introduced by sequence-based IBD inference), and directly evaluated how selection affects IBD-based demography estimation in P.f.. With high-quality IBD as input, we found that selection duration, strength, and selected allele frequency all affect IBD coverage and length distribution, and thus the IBD-based estimation of parasite Ne (IBDNe). The effects are further aggravated when reductions in Ne occur in a recent time frame. We found that removing IBD segments within the high-IBD region associated with selection and splitting the chromosome abolishes selection-induced bias in Ne estimation, especially for the most recent 30 generations. We also designed a model of multiple subpopulations with various migration parameters to test the effect of directional selection on IBD-based population structure estimation. We hypothesize that strong directional selection could increase within and inter-population IBD sharing and hide the underlying population structure. However, using filtered IBD segments could recalibrate the inference and thus allow identification of finer-scale, more accurate population structure in low transmission settings with active migration of drug-resistant parasites. Ongoing work will include genealogy inference, tree-based IBD finding, and recalibrated IBD-based estimation of P.f. demography using whole-genome sequencing data from field isolates.

Using Tree-Based Identity-By-Descent Segments to Evaluate the Effect of Directional Selection on the Estimation of Recent Effective Population Size and Population Structure in Plasmodium falciparum