Lassen Jan, Sørensen Morten Kargo, Madsen Per, Ducrocq Vincent
Department of Genetics and Biotechnology, Danish Institute of Agricultural Sciences, DK-8830 Tjele, Denmark.
Genet Sel Evol. 2007 Jul-Aug;39(4):353-67. doi: 10.1186/1297-9686-39-4-353. Epub 2007 Jul 6.
In a stochastic simulation study of a dairy cattle population three multitrait models for estimation of genetic parameters and prediction of breeding values were compared. The first model was an approximate multitrait model using a two-step procedure. The first step was a single trait model for all traits. The solutions for fixed effects from these analyses were subtracted from the phenotypes. A multitrait model only containing an overall mean, an additive genetic and a residual term was applied on these preadjusted data. The second model was similar to the first model, but the multitrait model also contained a year effect. The third model was a full multitrait model. Genetic trends for total merit and for the individual traits in the breeding goal were compared for the three scenarios to rank the models. The full multitrait model gave the highest genetic response, but was not significantly better than the approximate multitrait model including a year effect. The inclusion of a year effect into the second step of the approximate multitrait model significantly improved the genetic trend for total merit. In this study, estimation of genetic parameters for breeding value estimation using models corresponding to the ones used for prediction of breeding values increased the accuracy on the breeding values and thereby the genetic progress.
在一项对奶牛群体的随机模拟研究中，比较了三种用于估计遗传参数和预测育种值的多性状模型。第一个模型是使用两步法的近似多性状模型。第一步是针对所有性状的单性状模型。从这些分析中得出的固定效应解从表型中减去。在这些预先调整的数据上应用仅包含总体均值、加性遗传项和残差项的多性状模型。第二个模型与第一个模型相似，但多性状模型还包含年份效应。第三个模型是完整的多性状模型。比较了三种情况下育种目标中总价值和个体性状的遗传趋势，以对模型进行排名。完整的多性状模型给出了最高的遗传响应，但并不比包含年份效应的近似多性状模型显著更好。在近似多性状模型的第二步中纳入年份效应显著改善了总价值的遗传趋势。在本研究中，使用与用于预测育种值的模型相对应的模型来估计育种值的遗传参数，提高了育种值的准确性，从而提高了遗传进展。
