PSXII-1 - CNV and CNVR Detection from High-density SNP Genotyping in Holstein Cattle

Abstract: Copy number variation (CNV) represents an important source of genetic variation complementary to single nucleotide polymorphism (SNP) and provides valuable insights into the genetic architecture of complex traits in humans and livestock animals. However, many factors affect CNV detection, including sample size. Previous studies in Holstein cattle have only used hundreds of individuals. Therefore, in the present study, CNVs were identified from high-density genotype array data in 3,529 Holstein cows from 16 herds located in 7 U.S. states. CNVs were derived from 546,802 autosomal SNP markers with MAF > 0.05 and call rate > 0.90 using the intensity signals at each SNP position in a hidden Markov model implemented in PennCNV software. The log R ratio (LRR) values were corrected based on the guanine-cytosine content in the genomic regions located 500 kb upstream and downstream of each marker. After detection, CNVs presenting LRR ≤ 0.30, B allele frequency drift < 0.01, waviness factor ≤ 0.05, number of SNPs > 10 and length > 1 kb were merged into CNV regions (CNVR) using CNVRanger. A total of 1,881 non-redundant CNV events spanning the entire genome were identified in 1,771 cows, with a mean of 10 CNVs per individual and ranging from 1 to 27. The highest percentage of CNV calls was located on BTA7 (16%), while only 9 CNVs were identified on BTA9. The CNV calls were grouped into 685 CNVRs covering 3.07% of the bovine autosomal genome. The number of CNVRs with copy loss and gain were 356 and 310, and both types were observed in 19 regions. The chromosomal distribution of CNVR revealed that BTA7 harbors 38 CNVRs, while only 2 were detected on BTA28. These results will support further research on the contribution of CNVs and CNVRs to the genetic architecture of complex traits in Holstein cattle.