Data Processing
DNA sequencing was completed over four Illumina HiSeq runs with a total
of 2,258,652,346 reads, targeting at least 2 million reads per
individual (average reads per individual = 6,398,449). After cleaning,
mapping, and filtering the raw reads to have a quality level of at least
25 and no more than 5 SNPs per read, we had a total of 301,255,910
paired mapped reads (average per individual = 853,416) for the Stacks
analysis. Stacks identified 741,532 variable sites, which were then
filtered to a total of 7,274 variable sites (Supplemental Table 2).
These filtered, high-quality sites form the dataset used in all further
analysis unless otherwise noted.