Figure 1
From: Uninformative polymorphisms bias genome scans for signatures of selection
Informative and uninformative markers in genome scans. Two populations derived from the same ancestral population occupy ecologically distinct environments (white and gray boxes) at t0,. Circles represent an ecologically important QTL with two alleles under divergent selection; white and gray alleles are favored in the white and gray environment. Squares represent a neutral marker with two alleles (yellow and blue). The marker is tightly physically linked to the QTL. In A), both initial (t0) populations display a very low frequency for the blue marker allele. A novel adaptive QTL allele arising in the gray habitat will therefore likely be associated with the frequent yellow marker allele. When sampling the populations at t1, after a period of selection that has increased the frequency of the gray QTL allele in the gray environment, no signature of selection is visible at the marker locus because hitchhiking along with the favored QTL allele has not materially changed the allele frequency distribution at the marker (FST[22] approximates zero at both t0 and t1). In B), the initial conditions (t0) are as in A), except that the novel adaptive QTL allele happens to be linked to the rare blue marker allele. At t1, selection at the QTL will be visible at the marker (FST = 0.22) because the blue allele has hitchhiked to high frequency. In C), the initial (t0) allele frequency distribution at the marker is relatively even in both populations (FST = 0). At t1, the marker exhibits a clear signature of selection (FST = 0.13) because the yellow allele has increased in frequency by hitchhiking. In both B) and C) but not in A), we would consider the marker locus informative at t1 based on its minor allele frequency across both samples, and consider the marker for a genome scan for the signature of selection (see text).