Effect of errors in pedigree on the accuracy of estimated breeding value for carcass traits in Korean Hanwoo cattle

Abstract

Objective: This study evaluated the effect of pedigree errors (PEs) on the accuracy of estimated breeding value (EBV) and genetic gain for carcass traits in Korean Hanwoo cattle.Methods: The raw data set was based on the pedigree records of Korean Hanwoo cattle. The animals’ information was obtained using Hanwoo registration records from Korean animal improvement association database. The record comprised of 46,704 animals, where the number of the sires used was 1,298 and the dams were 38,366 animals. The traits considered were carcass weight (CWT), eye muscle area (EMA), back fat thickness (BFT), and marbling score (MS). Errors were introduced in the pedigree dataset through randomly assigning sires to all progenies. The error rates substituted were 5%, 10%, 20%, 30%, 40%, 50%, 60%, 70%, and 80%, respectively. A simulation was performed to produce a population of 1,650 animals from the pedigree data. A restricted maximum likelihood based animal model was applied to estimate the EBV, accuracy of the EBV, expected genetic gain, variance components, and heritability (h2) estimates for carcass traits. Correlation of the simulated data under PEs was also estimated using Pearson’s method.Results: The results showed that the carcass traits per slaughter year were not consistent. The average CWT, EMA, BFT, and MS were 342.60 kg, 78.76 cm<sup>2, 8.63 mm, and 3.31, respectively. When errors were introduced in the pedigree, the accuracy of EBV, genetic gain and h2 of carcass traits was reduced in this study. In addition, the correlation of the simulation was slightly affected under PEs.Conclusion: This study reveals the effect of PEs on the accuracy of EBV and genetic parameters for carcass traits, which provides valuable information for further study in Korean Hanwoo cattle.

Assessment of genomic prediction accuracy using different selection and evaluation approaches in a simulated Korean beef cattle population

“…The SampleBy function in the doBy package of the R software package (The R Foundation for Statistical Computing, Vienna, Austria) was used for creating PEs. Several error rates were substituted in the pedigree dataset for each generation, such as 10%, 20%, 30%, and 40%, respectively, as described previously by Oliehoek and Bijma [ 23 ] and adopted by Nwogwugwu et al [ 10 ].…”

“…However, the prediction accuracy of GEBVs can be influenced by several factors, such as methods of prediction [ 7 ], the training population (TP) size [ 6 ], the h2 of the trait [ 8 ] and the marker density [ 9 ]. An effect of errors in pedigree on the accuracy of the EBV has been reported in Korean Hanwoo cattle [ 10 ], but the use of genomic information could improve the accuracy of GEBVs.…”

“…Applying GS is a potential approach to improve the genetic gains in economically important traits. However, Onogi et al [ 17 ] pointed out that the use of ssGBLUP for genomic evaluations is still emerging in beef cattle due to the greater complexity of their records compared with those of other livestock species, such as the existence of pedigree errors (PEs) and fewer full or half-sib families [ 10 , 18 ]. On the other hand, a simulation study allows the testing of several theories, permitting an unravelling of the complex evolutionary patterns that are otherwise difficult to comprehend.…”

“…The SampleBy 143 function in the doBy package of the R software package (The R Foundation for Statistical Computing, 144 Vienna, Austria) was used for creating PEs. Several error rates were substituted in the pedigree dataset 145 for each generation, such as 10%, 20%, 30%, and 40%, respectively, as described previously by 146 Oliehoek and Bijma [23] and adopted by Nwogwugwu et al [10]. where y, µ, g, and e are the vectors of the phenotypes, overall mean, DGV and residual errors, 155 respectively, and Z is the incidence matrix that relates phenotypes to random marker effects (DGVs).…”

Accounting for Genetic Differences Among Unknown Parents in Bubalus bubalis: A Case Study From the Italian Mediterranean Buffalo

“…The problem of incomplete pedigree has existed for many years and continues to be one of the main issues in genetic evaluations. Several researchers have worked on possible statistical approaches in order to correct for the presence of gaps in the pedigree ( Peškovičová et al, 2004 ; Carneiro et al, 2017 ; Tonussi et al, 2017 ; Shiotsuki et al, 2018 ; Nwogwugwu et al, 2020 ; Macedo et al, 2020 ). The implementation of new technologies such as high-throughput single-nucleotide polymorphism (SNP) genotyping will certainly solve most of the problems linked to uncertain paternity but this is true only for individuals who are still alive or whose biological samples are available.…”