A Genomic Map of the Effects of Linked Selection in Drosophila

摘要: Natural selection at one site shapes patterns of genetic variation linked sites. Quantifying the effects "linked selection" on levels diversity is key to making reliable inference about demography, building a null model in scans for targets adaptation, and learning dynamics natural selection. Here, we introduce first method that jointly infers parameters distinct modes selection, notably background selective sweeps, from genome-wide data, functional annotations maps. The central idea calculate probability neutral polymorphic given local annotations, substitution patterns, recombination rates. Information then combined across sites samples using composite likelihood order estimate In addition parameter estimation, this approach yields map expected along genome. To illustrate utility our approach, apply it resequencing data 125 lines Drosophila melanogaster reliably predict 1Mb scale. Our results corroborate estimates high fraction beneficial substitutions proteins untranslated regions (UTR). They allow us distinguish between contribution sweeps other around amino acid uncover evidence pervasive (UTRs). further suggests substantial effect adaptation particular. More generally, demonstrate has had larger reducing increasing their variance D. than previously appreciated.