engleski

Estimation of genomic inbreeding in cattle: the impact of density

Runs of homozygosity (ROH) were recently proposed as genomic estimate of inbreeding that accounts for stochastic variation. In addition ROHs provide inference of more remote inbreeding than estimates obtained from pedigree. We analyzed impact of SNP chip density [Illumina BovineSNP50 Genotyping BeadChip(54k chip) and BovineHD Genotyping BeadChip (777k chip)] on the inbreeding estimates calculated for 115 Tyrolean Grey bulls. Inbreeding was calculated for whole pedigree (FpedT), for five generation pedigree(Fped5), and from five ROH lengths (k>1Mb, k>2Mb, k>4Mb, k>8Mb, k>16Mb) for both SNP chips (FROHk_54 & FROHk_777). ROH greater than 1Mb representing overall inbreeding of individual cover (on average) 7.7%, 15.5%, 5.4%, 6.2%, 6.8% and 7.7% of genome for FROH1_54_0, FROH1_54_1, FROH1_777_1, FROH1_777_1, FROH1_777_2 and FROH1_777_3, respectively. Pedigree inbreeding coefficients FpedT (2.2%) and Fped5 (1.4%) indicates much lower levels. FROHs from same lengths, regardless to the SNP chip density, were showing high correlations ranging from 0.90 to 0.99. Correlations of Fped and FROH were in range between 0.73 to 0.81. We conclude that correlations between pedigree and genomic inbreeding coefficients with respect to density were approximately the same. FROH estimates of inbreeding coefficients from 777k chip were lower. The total length of ROHs >1Mb is much shorter when considering 777k chip data compared to 54k chip data. Allowing one SNP in a ROH to be heterozyous for 777k chip data results in similar levels of inbreeding as derived from 54k chip data. Given the chance of genotyping errors, it seems advisable to allow one heterozyous SNP in a ROH.

Runs of homozygosity; SNP chip density; cattle

nije evidentirano