3.4. Genome wide test of selection
While we were able to detect signatures of selective sweeps in all three species, the patterns were strikingly different between G. calmariensis and the other two species (Fig. 2). In G.calmariensis , the outlier loci were concentrated in a handful of locations, which had strong signatures of selection large and were large (Fig. 3). In contrast, both G. tenella and G. pusillaexhibited signatures of selection that were generally weaker, with identified genomic regions that were more evenly distributed across their genomes (Fig. 2) and narrower in G. tenella and G. pusilla than in G. calmariensis (Fig. 3). To make a direct comparison with the HDMKPFR test that is focused upon the coding region of genes, we identified genes having a direct overlap of the sweep region with any part of a gene body, from the start to stop, and in between. This intersection of the outlier haplotypes with annotated genes revealed a total of 192, 115, and 154 genes in G.calmariensis, G. pusilla and G. tenella, respectively. When we included 10 kb on either side of each gene when detecting overlap with outlier haplotypes, a total of 216, 305, and 430 genes were intersected, suggesting that in the latter two species, there was more overlap with potential regulatory regions. We next sought to assess to what extent the identified candidate targets of positive selection from the genome wide analysis overlapped with our candidate immune genes. Unlike our results for the HDPRFMK test, the distribution of these outlier genes (with 10kb on either side) were roughly equally distributed among the three species (7, 6, and 6 genes in G.calmariensis, G. pusilla and G. tenella, respectively).