Fig. 3. The Alba locus in C. eurytheme.

(A) C. eurytheme specimens, depicting female phenotypes and genotypes. Alba is the dominant allele. (B) Results of the GWAS using data from 14 orange and 15 Alba wild-caught C. eurytheme females aligned against the Alba reference genome. Alternating gray-scale blocks in the Manhattan plot are colored by chromosome, ordered Z:31; the y axis represents negative log value Bonferroni-Holm–corrected false discovery rate [−log10(FDR_BH)] of each variant, the x axis is scaffold position, and the 0.05 and 0.01 percentile of the distribution is indicated with red and blue horizontal dashed lines, respectively. The second row is a close-up of scaffold 2 from chromosome 3, which contains the Alba locus. (C) Detailed view of the 58-kb region harboring the SNPs significantly associated with Alba. The gene model indicates the location of the BarH1 gene and its antisense reading frame (blue). Across this region, we show the read-mapping depth of whole-genome sequence data from an Alba (blue) and orange (orange) female from C. eurytheme (top row) and C. crocea (bottom row), respectively. The reads were aligned to the entire Alba C. eurytheme reference genome (filtered to MAPQ > 20 and proper pairs). Colored boxes highlight the Alba insertion (green), annotated repetitive content within the insertion (gray), and a region absent in orange individuals of both Colias species, which we refer to as the Alba candidate locus (pink). The gray region unique to Alba individuals in C. eurytheme is not unique to Alba individuals in C. crocea and is found to have elevated read depth coverage in all Eurasian species. The high variance in coverage to the left of the BarH1 locus is due to the high levels of repetitive content in this region, with reads mapping multiply and getting filtered out.