The use of runs of homozygosity for estimation of recent inbreeding in Holstein cattle

Controlling inbreeding in livestock populations is of great importance because excess relatedness among animals leads to a rapid loss of genetic variation and to adverse phenotypical effects associated with an inbreeding depression. Recent advances in genotyping technology have made it possible to study inbreeding at a molecular level by the analysis of genome-wide single nucleotide polymorphism panels. In this study, we used BovineSNP50 assay (Illumina) to estimate genomic inbreeding coefficient in 298 Holstein cattle by the analysis of the genome portion in runs of homozygosity (F-ROH) or using genomic relationship matrix (F-GRM), and compared this data with conventional pedigree-based inbreeding coefficients (F-PED). Weak or moderate Spearman's rank correlations were observed between F-ROH and F-PED which depended on the ROH length categories used for calculations and inclusion of animals with different number of complete generations registered in pedigrees. The highest correlations were observed when using ROH with lengths over 8 Mb (0.334). The correlations tended to increase as pedigree depth increased, and were the highest for animals with seven complete generations of pedigree data. F-GRM correlated poorly with pedigree-based estimates, which suggests that ROH-based inbreeding coefficients better reflect recent relatedness among animals.