School of Agriculture, Food and Ecosystem Sciences - Theses
Permanent URI for this collection
Search Results
1 - 1 of 1
-
ItemStructural variations in the ovine genome: their detection and association with phenotypic traits( 2013)Growth and meat yield traits are important to the New Zealand sheep industry. Genomic selection (GS) of these traits uses information from high density ovine SNPs to produce estimated breeding values (EBVs). The aim of GS is to account for loci directly influencing the trait (quantitative trait loci, QTL). This relies on the assumption that high density SNPs tag QTL effects via linkage disequilibrium (LD), however, it is unlikely that all QTL are adequately tagged by high density SNPs. Copy number variants (CNVs) are a type of genetic variant that may not be well tagged by SNPs and have been shown to be involved in phenotypic variation. To date, there has been little published work on CNVs in the sheep genome. While there is a well known example of a CNV affecting coat colour (agouti) in sheep, little is known of how CNVs affect phenotypic variation of production traits. The studies in this thesis employed multiple methods to identify CNVs in the sheep genome. Animals (including trios) were assayed on a Roche NimbleGen 2.1M CGH array. CNV calls from trios were used along with known false-positive calls to build a logistic regression to predict the probability calls from the 2.1M CGH array were correct. 3,488 autosomal CNVRs were identified. On a large scale, CNVRs were hard to accurately detect without using a combination of approaches. CNVRs were verified against CNVRs detected with the Roche NimbleGen 385K CGH array, Illumina OvineSNP50 BeadChip and Illumina HiSeq 2000 sequence data. Results of this work contribute a comprehensive resource of CNV regions to the literature on sheep CNVs. Given the importance of growth and meat yield traits in the New Zealand sheep industry, and the possibly unaccounted effects of CNVs on these traits, an association analysis was carried out with these traits and loci that potentially represent CNVs. Firstly, it was determined that EBVs produced by Sheep Improvement Limited (SIL) were appropriate to use as the dependent variable in the association analysis. Loci that potentially represent CNVs were SNPs from the Illumina OvineSNP50 BeadChip that were previously discarded from GS and genome wide association studies (GWAS) because they could not be genotyped. Reasons why these SNPs can’t be genotyped include the presence of the SNP in a CNV. Raw data from these SNPs were tested to determine if they were associated with the growth and meat yield traits. Seventeen associations, involving nine SNPs, were detected and validated in independent datasets. Two SNPs were in CNVRs detected using the CNV detection methods described above - one involved the agouti CNV. Raw data from this SNP was associated with ultrasonic eye muscle depth. Associations remained significant after fitting genotypes of flanking SNPs (from surrounding ~1Mb of sequence) used in GS and GWAS, suggesting that the effect of these associations are not accounted for in GS or GWAS. Including information from these SNPs in GS could improve the reliability of EBVs, contributing to genetic improvement of growth and meat yield traits in the New Zealand sheep industry.