Mao Xiaowei, Johansson Anna Maria, Sahana Goutam, Guldbrandtsen Bernt, De Koning Dirk-Jan
Department of Molecular Biology and Genetics, Center for Quantitative Genetics and Genomics, Aarhus University, PO Box 50, DK-8830, Tjele, Denmark; Department of Animal Breeding and Genetics, Swedish University of Agricultural Sciences, Box 7023, 75007 Uppsala, Sweden.
Department of Animal Breeding and Genetics, Swedish University of Agricultural Sciences, Box 7023, 75007 Uppsala, Sweden.
J Dairy Sci. 2016 Sep;99(9):7313-7318. doi: 10.3168/jds.2016-11160. Epub 2016 Aug 8.
Imputation is a cost-effective approach to augment marker data for genomic selection and genome-wide association studies. However, most imputation studies have focused on autosomes. Here, we assessed the imputation of markers on the X chromosome in Holstein cattle for nongenotyped animals and animals genotyped with low-density (Illumina BovineLD, Illumina Inc., San Diego, CA) chips, using animals genotyped with medium-density (Illumina BovineSNP50) chips. A total of 26,884 genotyped Holstein individuals genotyped with medium-density chips were used in this study. Imputation was carried out using FImpute V2.2. The following parameters were examined: treating the pseudoautosomal region as autosomal or as X specific, different sizes of reference groups, different male/female proportions in the reference group, and cumulated degree of relationship between the reference group and target group. The imputation accuracy of markers on the X chromosome was improved if the pseudoautosomal region was treated as autosomal. Increasing the proportion of females in the reference group improved the imputation accuracy for the X chromosome. Imputation for nongenotyped animals in general had lower accuracy compared with animals genotyped with the low-density single nucleotide polymorphism array. In addition, higher cumulative pedigree relationships between the reference group and the target animal led to higher imputation accuracy. In the future, better marker coverage of the X chromosome should be developed to facilitate genomic studies involving the X chromosome.
插补是一种经济高效的方法，可用于增加基因组选择和全基因组关联研究的标记数据。然而，大多数插补研究都集中在常染色体上。在此，我们使用中等密度（Illumina BovineSNP50）芯片进行基因分型的动物，评估了荷斯坦奶牛X染色体上标记对未进行基因分型的动物以及使用低密度（Illumina BovineLD，Illumina公司，加利福尼亚州圣地亚哥）芯片进行基因分型的动物的插补情况。本研究共使用了26,884头用中等密度芯片进行基因分型的荷斯坦个体。使用FImpute V2.2进行插补。研究了以下参数：将假常染色体区域视为常染色体或X特异性区域、不同大小的参考群体、参考群体中不同的雄性/雌性比例以及参考群体与目标群体之间的累积亲缘关系程度。如果将假常染色体区域视为常染色体，则X染色体上标记的插补准确性会提高。增加参考群体中雌性的比例可提高X染色体的插补准确性。一般来说，未进行基因分型的动物的插补准确性低于使用低密度单核苷酸多态性阵列进行基因分型的动物。此外，参考群体与目标动物之间较高的累积系谱关系导致更高的插补准确性。未来，应开发更好的X染色体标记覆盖范围，以促进涉及X染色体的基因组研究。
