Many studies have explored the ROH pattern and inbreeding depression at the genomic level in multiple cattle populations [39–41]. To our knowledge, our study is the first attempt to present the occurrence and distribution of ROH in commercial beef cattle population using high-density SNP arrays. Previous study suggested that high-density SNP arrays are more sensitive to determine the small segments, while Bovine SNP50K arrays may underestimate the number of fragments with length of 1–4 Mb [42]. In present study, our results showed the most abundance of ROH segments were between 0.5 and 1 Mb, which implied the power of identification of small ROH using high density SNP array.
We estimated inbreeding coefficients in the commercial beef cattle population using two methods including FROH and FHOM. In general, the correlation between FROH and FHOM varies across many studies. Previous study found the correlation range 0.78 to 0.85, indicating that FROH has a high correlation with the FHOM, thus FROH can be used as an accurate estimate of the proportion of IBD genomes [43]. In this study, the significant correlation (p < 2.2e-16) were observed between FROH and FHOM, which is consistent with previous report [44].
Many studies have shown significant differences in the total number and length of ROH in cattle, as well as the genetic relationship between individuals [7, 45]. The pattern of ROH count and length may indicate the differences of breed formation and recent breed management [7]. For instance, analysis of the British Isles breeds including Hereford, Guernsey, Angus and Jersey cattle displayed the highest sum of ROH events per animal compared other breeds, while African breeds displayed high variability in total number of ROH among breeds [7]. A recent study revealed that the average total length of ROH was 106 Mb and 371 Mb for Piedmontese cattle and Brown cattle, and the study also showed that ROHs identified in dairy breeds was longer and larger than beef and dual-purpose cattle. Consistent with previous findings, our results show ROH sizes range from 500 kb to several megabytes and also length and number of ROH are vary in commercial beef cattle population. Several studies suggest that short ROH reflect ancient inbreeding, and long ROH segments reveal recent inbreeding [31, 46, 47]. Therefore, our study revealed that most of ROH belong to the short and Medium, and this also indicate that the commercial population has low inbreeding level, which are agree with population history that this population have undergo hybrid process for recent selection.
Our study identified many consensuses of ROH across genome among population, and the distribution of ROH across genomic regions can imply functional effect for traits and help to identify functional pathways affected by inbreeding. In this study, we totally detected 26 regions with ROH frequency exceeding 10% among population, these regions were overlapped with 16 QTLs related to important traits including weight gain, calving difficulty, Gestation length, Lean meat yield and Stillbirth, which may potentially imply their impacts on reduction in performance for important trait [38].
Many previous studies have been investigated to detect the regions of homozygosity and their impact on complex traits in human [16, 19, 20, 48]. For instance, several methods have been utilized to explore the association between ROH region and complex disease [14, 45, 49, 50]. In current study, we proposed the region-based and loci-based approach to investigate the candidate the ROH region and loci associated with important traits in cattle. These approaches can also be extended to identify the loci of ROH related to important traits in other farm animals.
In present study, we firstly carried the association test based on ROH region. Totally, we obtained 280 regions by merging all individual ROH events among population, and then we utilized proportion of ROH coverage for each region among individual as variable. Using linear model, we identified 16 significant ROH regions, we found 371 candidate genes been located in nonredundant regions in the studied population. Based on the significant region, we obtained 153 genes, while other region with 218 genes. Since many genes were identified for the association analysis based on ROH region, therefore, it is difficult to pinpoint the region of ROH associated with studied trait.
Furthermore, we performed loci-based association to precisely locate the region of ROH for important trait. Totally, we found 1631 non redundant loci and 67 candidate genes for 8 traits. For the fat coverage, we found only one gene (EBF2) within 37 loci for fat coverage, this gene was found related to fat formation [51, 52]. Two genes (ODF1 and UBR5) with 4 loci and 23 loci were detected for slaughter Weight. One study found UBR5 gene is involved with glucose-dependent degradation of PEPCK1 [53]. For backfat thickness, 103 candidate loci within 9 candidate genes were identified in current study. However, only one gene (SLC20A2) was associated with carcass trait [54]. For carcass length, 188 loci and 31 candidate genes were identified. Among them, there most promising candidate genes including SH3BGRL2, HMGA1 and ACSL1 have been reported from previous report. For instance, SH3BGRL2 gene had a significant effect on fatty acid metabolism [55], and HMGA1 gene was related to the growth, fertility and lean meat content in commercial pigs [56], and ACSL1 gene was a candidate gene for the function of fatty acid composition in bovine skeletal muscle [57]. For the body height, we found 24 candidate genes embedded with 119 loci, of these gene, HMGA1 and ACSL1 have been implied that can promote the growth and development of animal body [57, 58].