Identification of genomic regions under selection
We used VCFtools v0.1.16 (Danecek et al., 2011) to calculate Tajima’s D
in 50 kb non-overlapping windows and nucleotide diversity (π) and
population-differentiation statistic (F ST) for
the two populations both in 50 kb sliding windows with 10 kb steps. To
identify regions of the genome that may have been subject to selection
associated with sperm variation of tree sparrow living in BY, we
detected the regions with both high divergence and low diversity,
characteristic of recent selective sweeps. We only selected the top 1%
of outliers with low diversity in BY population estimated by the
pairwise differences (πLJX/πBY)
converted to Z scores and high divergence estimated byF ST.