Reducing bias in the dairy cattle single-step genomic evaluation by ignoring bulls without progeny
Reducing bias in the dairy cattle single-step genomic evaluation by ignoring bulls without progeny
复制标题
DOI：
10.1111/jbg.12318
复制
发表时间：
2018-04-01
影响因子：
2.6
通讯作者：
Mantysaari, E. A.
中科院分区：
文献类型：
--
作者：
Koivula, M.;Stranden, I.;Mantysaari, E. A.
基因分型动物的数量迅速增加，为基因组评估带来了计算挑战。在动物模型BLUP中，没有后代和表型的候选动物不会为评估提供信息，可以被丢弃。理论上，没有后代的基因型候选动物可以将信息带入单步BLUP（ssGBLUP），并影响其他育种值的估计。我们研究了从ssGBLUP中包含或排除淘汰公牛小牛的基因组信息对基因组育种值（GEBV）的影响。特别地，研究了具有女儿的基因型公牛的GEBV和被选入AI以进行后代测试的年轻公牛（测试公牛）的GEBV。使用Nordic试验日（TD）模型和Nordic Red奶牛的TD数据计算ssGBLUP评价。结果表明，剔除公牛小牛的基因组信息不影响后代测试参考动物的GEBV，但如果在TD ssGBLUP中使用剔除公牛小牛的基因型，则测试公牛的遗传趋势显著高于剔除剔除公牛小牛的基因组信息的情况。似乎通过丢弃剔除无后代公牛犊的基因组信息，降低了测试公牛的GEBV的向上偏差。
The number of genotyped animals has increased rapidly creating computational challenges for genomic evaluation. In animal model BLUP, candidate animals without progeny and phenotype do not contribute information to the evaluation and can be discarded. In theory, genotyped candidate animal without progeny can bring information into single-step BLUP (ssGBLUP) and affect the estimation of other breeding values. We studied the effect of including or excluding genomic information of culled bull calves on genomic breeding values (GEBV) from ssGBLUP. In particular, GEBVs of genotyped bulls with daughters and GEBVs of young bulls selected into AI to be progeny tested (test bulls) were studied. The ssGBLUP evaluation was computed using Nordic test day (TD) model and TD data for the Nordic Red Dairy Cattle. The results indicate that genomic information of culled bull calves does not affect the GEBVs of progeny tested reference animals, but if genotypes of the culled bulls are used in the TD ssGBLUP, the genetic trend in the test bulls is considerably higher compared to the situation when genomic information of the culled bull calves is excluded. It seems that by discarding genomic information of culled bull calves without progeny, upward bias of GEBVs of test bulls is reduced.