Novel algorithm to determine breed composition based on haplotypes

Breed composition of animals has traditionally been inferred from pedigree records; however, with the availability of genotypes, molecular marker information could be used instead. Moreover, breed composition can deviate from parental expectations due to Mendelian sampling and recombination. In this context, molecular markers, particularly single-nucleotide polymorphisms (SNPs), offer a more accurate way for estimating breed composition. The accuracy of most methods for determining breed composition from genomic information depends on genotypes from purebred individuals within the reference population. Missing data for one or more purebred populations can introduce bias into breed composition predictions. The novel algorithm evaluated in this study uses haplotype contributions and haplotype ages within the reference population. Haplotype frequency within a breed is dynamic and may change substantially due to selection and drift. Therefore, individuals from earlier generations may have a different genetic architecture compared to more recent individuals, even if they belong to the same pure breed. The novel haplotype-based algorithm estimates both breed purity and breed composition. The objective of this study was to evaluate the novel haplotype-based algorithm implemented in snp1101 for estimating breed composition in purebred and crossbred animals, and to compare its performance with the Admixture software. Eight purebred breeds, including Belclare, Beltex, Charollais, Cheviot, Galway, Suffolk, Texel, and Vendeen, and nine types of crosses with known ancestry were genotyped using the Illumina OvineSNP50 BeadChip. Quality control excluded samples and SNPs with call rates below 95%, SNPs with MAF highlighting its potential to identify outcross animals within pure breeds. Overall, the haplotype-based algorithm implemented in snp1101 showed strong potential. Performance evaluation based on pedigree-derived purity expectations should be interpreted with caution, as pedigree-derived expectations may not fully reflect the underlying genomic composition.