Skip to Main content Skip to Navigation

Combined purebred and crossbred information for genomic evaluation in pig

Abstract : This PhD thesis has two aims: first, apply single-step genomic evaluation method for purebred and crossbred performances in different scenarios with data records and genotypes in Danish Landrace, Yorkshire and F1 crossbred pig populations; second, investigate the impact of non-additive genetic effects on genomic evaluation for crossbred performance. In chapter 2, performances of genotype imputation in low density SNP-panels were compared in both purebred and crossbred populations. Imputation for crossbreds worked as well as for purebreds if both parental breeds were included in the reference population. In chapter 3, the single-step GBLUP method was applied to a combined purebred and crossbred dataset, focusing on evaluating genetic ability for crossbred performance of total number of piglets born (TNB). Additive genetic effects in crossbred animals were split into two breed-specific gametic effects. The analysis confirmed the existence of a moderate, positive genetic correlation between purebred and crossbred performances for TNB. Models with genomic information, especially from crossbred animals, improved model-based reliabilities for crossbred performance of purebred boars and also improve predictive abilities on crossbred animals in a validation population. This method requires tracing the breed origin of crossbred alleles, which may be inconvenient. Therefore, in chapter 4, this dataset was reanalysed using a single relationship matrix with metafounders to relate all the involved animals in the three populations. This method did not need tracing the breed origin of crossbred alleles. Estimates of genetic parameters were similar to those in chapter 3 and the predictive abilities for crossbred performance were at least as good as in chapter 3. Both chapters 3 and 4 indicate that the single-step method for combined purebred and crossbred performances is applicable for genomic evaluation. In chapter 5, genomic evaluation using a model including dominance effects and inbreeding depression was investigated for genotyped animals in GBLUP context. The estimated correlations between allele substitution effects of markers for different breeds were reported for the first time. Results indicated that the accuracies of predictions were not improved by including dominance effects in the model, but inclusion of genomic inbreeding depression effects did improve the performance of prediction.
Document type :
Complete list of metadata

Cited literature [202 references]  Display  Hide  Download
Contributor : ABES STAR :  Contact
Submitted on : Friday, June 26, 2020 - 3:13:12 PM
Last modification on : Friday, August 5, 2022 - 2:38:10 PM


Version validated by the jury (STAR)


  • HAL Id : tel-02882220, version 1


Tao Xiang. Combined purebred and crossbred information for genomic evaluation in pig. Zootechny. Institut agronomique, vétérinaire et forestier de France; Aarhus universitet (Danemark), 2017. English. ⟨NNT : 2017IAVF0014⟩. ⟨tel-02882220⟩



Record views


Files downloads