Abstract:
Genotyping panels using single nucleotide polymorphism (SNP) markers have superseded the use of microsatellites for parentage testing and other applications. High-density SNP arrays enable accurate parentage assignment but they are not always practical for routine application. The major constraints include high cost coupled with the time spent in analyzing the results. The International Society for Animal Genetics (ISAG) recommends a low-density SNP panel consisting of 200 genomic markers for parentage testing globally. However, SNPs specific to Sanga cattle breeds were not included in constructing the panel, and its utility for use has not been verified in Sanga cattle. This thesis, therefore, firstly tested the usefulness of these ISAG markers for parentage testing in Sanga cattle with verified parentage. The results demonstrated a poor performance of the ISAG panel in both Bonsmara (BON) and Drakensberger (DRB) for parentage verification, with false negatives ranging from 23.4% (BON) to 33% (DRB). This implied that relying on the ISAG panel alone may cause incorrect exclusions and, at times, be unable to determine parentages when closely related candidate parents are considered. The second objective of the thesis was to quantify the accuracy of parentage recording but also the detection of individuals exhibiting hemizygous deletions using 91 185 autosomal SNPs. On average, 8.5% to 10.1% of parentage errors were detected in the recorded BON and DRB pedigree, respectively indicating recording errors on farms. The discovery of the parent-progeny pairs based on SNP genotypes was possible for 69 relationships that had not actually been recorded. Eleven suggestive regions of hemizygosity were detected on 10 chromosomes (3, 6, 7, 9, 10, 11, 17, 24, 25 and 28) in animals mostly used for breeding which included 6 (BON) and 9 (DRB) bulls and as well 10 (BON) and 8 (DRB) cows. Finally, low-density genotype SNP panels for parentage testing consisting of 200 markers were developed. SNPs were chosen to be informative both within and across breeds. The methods to select informative SNPs considered high minor allele frequency (MAF), good clustering quality, and high call rates. SNPs were pruned to reduce the linkage disequilibrium among markers with a minimum distance of one Mb apart. All the panels selected were tested per breed. On average, the genotype panel with SNPs selected across the breeds had a lower MAF of 0.40 compared to 0.48 (DRB) and 0.49 (BON) selected within breeds. SNP markers selected within breed were more accurate at parentage testing with no false negative whereas 4.2% false negatives were observed in the BON using the multi-breed panel. The methods and results presented in this thesis can be used in the construction of parentage SNP panels to provide parentage verification which will contribute to quality control in breeding systems and thus accelerate genetic improvement.
