Genetic Parameters for a Weighted Analysis of Survivability in Dairy Cattle

The genetic parameters for the survival of Holstein cows, analysed in nine consecutive time periods during the first three calving intervals, were estimated. The earlier the animals are culled, the more they are informationally underestimated. This undervaluing can be remedied by using a weighted analysis that balances the amount of information. If the method of estimating breeding values changes, the genetic parameters will also change. The Holstein cattle dataset from 2005 to 2017 used in this study included 1,813,636 survival records from 298,290 cows. The pedigree with three generations of ancestors included 660,476 individuals. Linear repeatability models estimated genetic parameters for overall and functional survivability. Due to weights, heritability increased from 0.013 to 0.057. Repeatability with weights was 0.505. The standard deviations of breeding values were 1.75 and 2.18 without weights and 6.04 and 6.20 with weights. Including weights in the calculation increased the additive variance proportion and the breeding values’ reliabilities. We conclude that the main contribution of the weighted method we have presented is to compensate for the lack of records in culled individuals with a positive impact on the reliability of the breeding value.

The prediction accuracies of linear-type traits in Czech Holstein cattle when using ssGBLUP or wssGBLUP

The aim of this study was to assess the contribution of the weighted single-step genomic best linear unbiased prediction (wssGBLUP) method compared to the single-step genomic best linear unbiased prediction (ssGBLUP) method for genomic evaluation of 25 linear-type traits in the Czech Holstein cattle population. The nationwide database of linear-type traits with 6,99,681 records combined with deregressed proofs from Interbull (MACE method) was used as the input data. Genomic breeding values (GEBVs) were predicted based on these phenotypes using ssGBLUP and wssGBLUP methods using the BLUPF90 software. The bull validation test was employed which was based on comparing GEBVs of young bulls (N = 334) with no progeny in 2016. A minimum of 50 daughters with their own performance in 2020 was chosen to verify the contribution to the GEBV prediction, GEBV reliability, validation reliabilities (R2), and regression coefficients (b1). The results showed that the differences between the two methods were negligible. The low benefit of wssGBLUP may be due to the inclusion of a small number of SNPs; therefore, most predictions rely on polygenic relationships between animals. Nevertheless, the benefits of wssGBLUP analysis should be assessed with respect to specific population structures and given traits.

Approximation of reliabilities for random-regression single-step genomic best linear unbiased predictor models

Random-regression models (RRM) are used in national genetic evaluations for longitudinal traits. The outputs of RRM are an index based on random-regression coefficients and its reliability. The reliabilities are obtained from the inverse of the coefficient matrix of mixed model equations (MME). The reliabilities must be approximated for large datasets because it is impossible to invert the MME. There is no extensive literature on methods to approximate the reliabilities of RRM when genomic information is included by single-step GBLUP. We developed an algorithm to approximate such reliabilities. Our method combines the reliability of the index without genomic information with the reliability of a GBLUP model in terms of effective record contributions. We tested our algorithm in the 3-lactation model for milk yield from the Czech Republic. The data had 30 million test-day records, 2.5 million animals in the pedigree, and 54,000 genotyped animals. The correlation between our approximation and the reliabilities obtained from the inversion of the MME was 0.98, and the slope and intercept of the regression were 0.91 and 0.02, respectively. The elapsed time to approximate the reliabilities for the Czech data was 21 min.