Association And Prediction Of Traits Based On Genomic Homozygous Segments In Beef Cattle

Runs of homozygosity（ROH）are continuous homozygous regions that generally exist in the DNA sequence of diploid organisms.Several studies have suggested that an individual inherits the chromosomal fragment from both parents,resulting in homozygous segments in the offspring’s genome.Identification of ROH patterns in population genomes can assess population inbreeding levels,refer population history,identify candidate genes for economic traits,and evaluate genetic diversity.In this study,we detected and analyzed ROH of Chinese Simmental beef cattle using Bovine 770K SNP array,estimated the correlation between ROH in the genome and economically important traits,and constructed predictive variables based on ROH to evaluate the accuracy of genetic selection on these traits.Our main research contents are as follows:（1）A total of 116953 ROH segments were detected in 1181 Chinese Simmental beef cattle,with an average of 99.03 ROH segments per individual,and the ROH segments average length of individual was 117.29 Mb.Within the population,the number of ROH detected by different individuals varies greatly（27-481）.In addition,we observed the number of ROH varied across autosomes,with the highest number was found on BTA1（9,176）and the minimum on BTA25（1,187）.The distribution of ROH segments with different lengths was also different in the population.The short ROH（0.5-1 Mb）accounted for the largest proportion of the total ROH number（69.60%）,while the long ROH（>5 Mb）with the smallest percentage（2.11%）.（2）In this study,a total of 8195 consentient ROH were detected across the genome in Chinese Simmental beef cattle.The percentage of consentient ROH enrichment on chromosomes was different,and we detected a total of 42 consentient ROH with frequencies over 20%.We found 17 candidate genes in 42 consentient ROH.These genes include DNAJC18,ECSCR,MATR3,PAIP2,SNHG4,SPATA24,TMEM173,UBE2D2,KED,DCAF16,LCORL,NCAPG.Meanwhile,QTLs were found to correlate with carcass weight,calving index,body weight,rump width,average daily gain and shear force.（3）We performed the correlation analyses between total length of ROH for each animal and adjusted phenotype including net meat weight（NMW）,carcass weight（CW）,average daily gain（ADG）and live weight（LW）,the Pearson’s correlation coefficients were 0.02,0.04,0.04 and 0.05,respectively.We also used short,medium,and long ROH for correlation analysis with four traits.Our results showed that the correlation coefficients were 0.13,0.15,0.14,0.15 for the total length of short ROH,and the correlation coefficients were-0.06,-0.054,-0.046,-0.049 for the total length of long ROH.We evaluated the significance of each ROH interval using Fisher’s exact test and obtained 164 ROH regions that were significantly associated with phenotypic traits.（4）Further association analysis using a mixed linear model revealed that 12 regions were significantly associated with production traits（P<0.05）.Totally,we observed 16 genes related to production traits,which may play an important role in promoting growth and metabolic functions.In addition,we found one region located at（chr1:95450372-95517810）for average daily gain and live weight,which was largely overlapped with ECT2.We also observed one ROH region on BTA9（chr9:38706395-38924566）with three genes（FAM229B,LAMA4 and TUBE1）associated with CW and NMW.（5）We constructed G matrix and performed the genomic prediction for six traits（NMW,CW,ADG,LW,CL and CD）using GBLUP approach.Based on five-fold cross validation,we compared the accuracy of prediction breeding values between the G_ROH based on ROH in the genome and G_IBS based on SNP in the genome.We found that the accuracy of genomic prediction for six traits were 0.2156,0.2225,0.2423,0.1881,0.1571 and 0.0851,while the prediction accuracies were 0.2618,0.2809,0.3179,0.2635,0.2047 and 0.1107 based on SNPs.Our results showed that the prediction accuracy of G_ROHbased on the genomic ROH was low.