Using genomic data in sheep breeding programmes

Abstract

There are 1.3 billion sheep in the world that provide milk, meat and wool to meet market demands for these products. In order to keep up with the increasing demand, sheep production must become more efficient whilst also considering the importance of animal welfare. The most sustainable route to achieve that is by improving both the environment within which animals are reared and the inherent capacity of the animals. The latter is achieved with holistic breeding programmes aiming to increase productivity while improving biological functions of the animals.

This thesis investigates the potential of adding genomic information (Single Nucleotide Polymorphisms, SNP) into the sheep breeding programme. Chapter two examines the potential of correcting the pedigree using both conventional (opposing homozygote) and unconventional (with genomic relationship matrix, GRM) method for looking into errors in the pedigree. The results provide evidence that parentage errors can be identified and resolved even when there is no genotypic information coming directly from the sire and/or dam, under the assumption that there is enough information coming from grandparents, aunts/uncles, or siblings. Chapter three dissects the genomic architecture of the Texel sheep population and examines the impact on changes in accuracy of breeding values, following the inclusion of genomic information into the Best Linear Unbiased Predictor (BLUP) analysis. The results provide evidence that the accuracy of prediction is higher while using the genomic approach, which is especially important for traits that are the sex-limited, measured late in life or have low heritability. Chapter four investigates the genetic background of health and production traits using the available phenotypic and SNP data. Genome-Wide Association Studies (GWAS) on new data pertaining to footrot and mastitis, as well as production traits identified informative markers that could be used in the BLUP model to increase the accuracies of predicting breeding values. The majority of the results provide evidence that all traits studied are highly polygenic.

The information presented in this thesis demonstrates that utilising SNP data in sheep breeding programmes can improve the quality of the data (pedigree) and increase the accuracy of predicting breeding values. Incorporating such genomic information into current breeding programmes for sheep will lead to improvements in accuracy, thereby reducing the risks associated with making selection decisions, which can be highly beneficial for the breeders of these animals. Due to the complexity of the genetic architecture of these traits, additional research to further fine-map the genomic areas highlighted in this study would potentially improve our understanding of the most informative SNP markers for the traits studied here.