1.29 1.47 1.52 1.36 1.30 1.26 1.28 1.40 1.63 1.45 1.28 1.40 1.81 1.43 1.35 1.25 1.52 1.50 1.37 1.27 2.04 1.23 1.44 1.47 1.37 1.50 1.46 1.43 1.q-value 0.033 0.038 0.043 0.006 0.010 0.017 0.020 0.021 0.028 0.031 0.036 0.007 0.005 0.023 0.034 0.041 0.040 0.018 0.001 0.011 0.047 0.014 0.003 0.012 0.025 0.045 0.004 0.008 0.015 0.050 0.000 0.048 0.010 0.009 0.026 0.006 0.008 0.013 0.N SNPsGuti rez-Gil et al.8 ???????52.3?2.5 Mb ?52.3?2.5 Mb ??18.9?9.3 MbFariello et al.7 ???????51.4?3.4 Mb ?51.4?3.4 Mb ??151.4?56.9 Mb ??35.9?8.3Mb ??????24.02?4.91 ??????6.3?3.6 ???30.4?5.09 ???OAR3_164170826.1 s38388.1 s19983.153.4?54.5 Mb ??OAR6_44123475_X.1 OAR7_30772408.1 s68972.1 s11241.1 OAR7_1827930.1 OAR9_14653377.1 OAR10_90168545.1 OAR10_23129120.1 OAR10_79676247.1 s34065.1 s56762.1 s38696.1 s05603.1 s19740.1 OAR14_9498278.1 s45350.1 OAR17_33487124.1 OAR17_23200636.1 s18836.1 OAR19_33355170.1 s06827.1 EPZ004777 solubility s67158.1 OAR25_23589759.39.3?9.5 Mb ?????????????16.6?0.6 Mb ???????9 10114 15 17 19 241 1 2 2 1Table 1. Outlier SNPs found with the FST-based method implemented in BayeScan. In the two columns at the right part of the table, we show evidence of positional concordance with CEP-37440 web selective sweeps detected with an FST-outlier approach and other methods8 as well as with the FLK and hapFLK metrics7. CHR = chromosome, N SNPs = Number of outlier SNPs.coincident selective sweep in the 37?8 Mb interval of Oar6 was previously found by analysing a set of Alpine ovine breeds with the hapFLK statistic2. As shown in Table 2, there was a substantial positional concordance between the set of selective sweeps detected with hapFLK and those described in previous reports7,8. This level of coincidence also exceeded what would be expected just by chance (bootstrapped P-value < 0.05) based on a circular permutation test. The set of Spanish populations analysed in our study is considerably different to those employed by other authors i.e.10 European breeds (2 of them with a Spanish origin)8, and 29 international breeds (only 1 Spanish breed)7. Although drift and migration can generate local signatures that can be confounded with those produced by selection, in principle we do not expect distantly related sets of populations to share such demographic signals. Moreover, the coincident FST-outlier signals found by us and others2,7,8 suggest that at least part of the selective sweeps detected with BayeScan are true positives (despite the fact that they were not detected with hapFLK).Three selective sweeps are consistently detected with BayeScan and hapFLK. When we consid-ered the BayeScan data set and the selective sweeps detected with hapFLK that are significant at the nominal levelScientific RepoRts | 6:27296 | DOI: 10.1038/srepwww.nature.com/scientificreports/Figure 3. Whole-genome scan for selective sweeps. The two analyses were based on either the FST-outlier method implemented in BayeScan (a) or the hapFLK statistic (b). In the BayeScan analysis, the red and blue lines indicate the thresholds of significance set at 0.05 and 0.01 after correction for multiple testing (q-values), respectively. In the hapFLK analysis, the red and blue lines indicate the thresholds of significance set at 0.05 before (nominal P-value) and after (q-value) correction for multiple testing, respectively. Genomic coordinates and statistical significance (-log10 P-values) are plotted in the x- and y-axis, respectively.CHR 2 3Reg (Mb) 82.8?7.7 150.5?54.2 46.5?9.1 4.3?9.9 52.3?2.6 64.5?4.3 112.1?15.6 29.1?9.3.1.29 1.47 1.52 1.36 1.30 1.26 1.28 1.40 1.63 1.45 1.28 1.40 1.81 1.43 1.35 1.25 1.52 1.50 1.37 1.27 2.04 1.23 1.44 1.47 1.37 1.50 1.46 1.43 1.q-value 0.033 0.038 0.043 0.006 0.010 0.017 0.020 0.021 0.028 0.031 0.036 0.007 0.005 0.023 0.034 0.041 0.040 0.018 0.001 0.011 0.047 0.014 0.003 0.012 0.025 0.045 0.004 0.008 0.015 0.050 0.000 0.048 0.010 0.009 0.026 0.006 0.008 0.013 0.N SNPsGuti rez-Gil et al.8 ???????52.3?2.5 Mb ?52.3?2.5 Mb ??18.9?9.3 MbFariello et al.7 ???????51.4?3.4 Mb ?51.4?3.4 Mb ??151.4?56.9 Mb ??35.9?8.3Mb ??????24.02?4.91 ??????6.3?3.6 ???30.4?5.09 ???OAR3_164170826.1 s38388.1 s19983.153.4?54.5 Mb ??OAR6_44123475_X.1 OAR7_30772408.1 s68972.1 s11241.1 OAR7_1827930.1 OAR9_14653377.1 OAR10_90168545.1 OAR10_23129120.1 OAR10_79676247.1 s34065.1 s56762.1 s38696.1 s05603.1 s19740.1 OAR14_9498278.1 s45350.1 OAR17_33487124.1 OAR17_23200636.1 s18836.1 OAR19_33355170.1 s06827.1 s67158.1 OAR25_23589759.39.3?9.5 Mb ?????????????16.6?0.6 Mb ???????9 10114 15 17 19 241 1 2 2 1Table 1. Outlier SNPs found with the FST-based method implemented in BayeScan. In the two columns at the right part of the table, we show evidence of positional concordance with selective sweeps detected with an FST-outlier approach and other methods8 as well as with the FLK and hapFLK metrics7. CHR = chromosome, N SNPs = Number of outlier SNPs.coincident selective sweep in the 37?8 Mb interval of Oar6 was previously found by analysing a set of Alpine ovine breeds with the hapFLK statistic2. As shown in Table 2, there was a substantial positional concordance between the set of selective sweeps detected with hapFLK and those described in previous reports7,8. This level of coincidence also exceeded what would be expected just by chance (bootstrapped P-value < 0.05) based on a circular permutation test. The set of Spanish populations analysed in our study is considerably different to those employed by other authors i.e.10 European breeds (2 of them with a Spanish origin)8, and 29 international breeds (only 1 Spanish breed)7. Although drift and migration can generate local signatures that can be confounded with those produced by selection, in principle we do not expect distantly related sets of populations to share such demographic signals. Moreover, the coincident FST-outlier signals found by us and others2,7,8 suggest that at least part of the selective sweeps detected with BayeScan are true positives (despite the fact that they were not detected with hapFLK).Three selective sweeps are consistently detected with BayeScan and hapFLK. When we consid-ered the BayeScan data set and the selective sweeps detected with hapFLK that are significant at the nominal levelScientific RepoRts | 6:27296 | DOI: 10.1038/srepwww.nature.com/scientificreports/Figure 3. Whole-genome scan for selective sweeps. The two analyses were based on either the FST-outlier method implemented in BayeScan (a) or the hapFLK statistic (b). In the BayeScan analysis, the red and blue lines indicate the thresholds of significance set at 0.05 and 0.01 after correction for multiple testing (q-values), respectively. In the hapFLK analysis, the red and blue lines indicate the thresholds of significance set at 0.05 before (nominal P-value) and after (q-value) correction for multiple testing, respectively. Genomic coordinates and statistical significance (-log10 P-values) are plotted in the x- and y-axis, respectively.CHR 2 3Reg (Mb) 82.8?7.7 150.5?54.2 46.5?9.1 4.3?9.9 52.3?2.6 64.5?4.3 112.1?15.6 29.1?9.3.
