Effects of Sample Size and Single- vs. Multiple-Breed Analyses on Characterizing Runs of Homozygosity in Dairy Cattle

The understanding and management of inbreeding is an ongoing challenge in the dairy industry. Runs of homozygosity (ROH) can provide more accurate estimates of the realized proportion of the genome shared between individuals. The objective of this thesis was to characterize ROH in dairy cattle breeds using different ROH detection methods. Data was prepared using PLINK and R, and BCFtools was the used to detect ROH. The R packages detectRUNS and pedigree were then used to characterize ROH and to calculate pedigree-based inbreeding respectively. SNP inclusion in single- and multiple-breed analyses impacted the degree of breed differences observed. Sample size was shown to affect the accuracy of ROH characterization. The results from this study indicate ROH could potentially be used manage inbreeding levels. Future research will aim to develop a method to minimize breed bias in multiple-breed analyses and to understand breed differences.