VolcanoFinder: genomic scans for adaptive introgression

AbstractRecent research shows that introgression between closely-related species is an important source of adaptive alleles for a wide range of taxa. Typically, detection of adaptive introgression from genomic data relies on comparative analyses that require sequence data from both the recipient and the donor species. However, in many cases, the donor is unknown or the data is not currently available. Here, we introduce a genome-scan method—VolcanoFinder—to detect recent events of adaptive introgression using polymorphism data from the recipient species only.VolcanoFinderdetects adaptive introgression sweeps from the pattern of excess intermediate-frequency polymorphism they produce in the flanking region of the genome, a pattern which appears as a volcano-shape in pairwise genetic diversity.Using coalescent theory, we derive analytical predictions for these patterns. Based on these results, we develop a composite-likelihood test to detect signatures of adaptive introgression relative to the genomic background. Simulation results show thatVolcanoFinderhas high statistical power to detect these signatures, even for older sweeps and for soft sweeps initiated by multiple migrant haplotypes. Finally, we implementVolcanoFinderto detect archaic introgression in European and sub-Saharan African human populations, and uncovered interesting candidates in both populations, such asTSHRin Europeans andTCHH-RPTNin Africans. We discuss their biological implications and provide guidelines for identifying and circumventing artifactual signals during empirical applications ofVolcanoFinder.Author summaryThe process by which beneficial alleles are introduced into a species from a closely-related species is termed adaptive introgression. We present an analytically-tractable model for the effects of adaptive introgression on non-adaptive genetic variation in the genomic region surrounding the beneficial allele. The result we describe is a characteristic volcano-shaped pattern of increased variability that arises around the positively-selected site, and we introduce an open-source methodVolcanoFinderto detect this signal in genomic data. Importantly,VolcanoFinderis a population-genetic likelihood-based approach, rather than a comparative-genomic approach, and can therefore probe genomic variation data from a single population for footprints of adaptive introgression, even froma prioriunknown and possibly extinct donor species.