High Resolution Copy Number Variable Regions in Brown Swiss Dairy Cattle and their Value as Markers

Recent studies have identified copy number variations (CNVs) that are functionally significant within mammalian genomes. This study presents a high resolution map of CNV regions (CNVRs) in Brown Swiss dairy cattle and characterizes identified CNVs as markers for quantitative and population genetic analysis. After careful quality filtering, CNVs were called in a set of 164 proven sires of the Brown Swiss breed with PennCNV and CNAM (Goldenhelix) at each of 735,238 loci (Illumina’s bovine HD SNP chip) anchored to bovine autosomes on the UMD3.1 assembly. Overlapping CNVs were summarized at the population level into CNVRs. PennCNV identified 2,377 CNVRs comprising 1,162 and 1,131 gain and loss events, respectively and 84 regions of complex nature. CNAM identified 370 CNVRs comprising 23 and 240 gain and loss events, respectively, and 107 comprising both. The concordance for low frequency and short CNV calls between methods was low, emphasizing the need to establish consensus calls employing several algorithms, to avoid false positives. We therefore defined 203 consensus CNVRs as the intersection of CNVRs identified by both algorithms. These covered 14.86 Mb (0.59%) of the autosome. Consensus CNVRs are enriched for protein coding-, pseudo- and retroposed genes and biological functions involved in immunity. We then used total allele counts called by genoCN in consensus CNVRs and phased them with CNVphaserPro to disentangle the alleles sitting on each chromosome. Characterisation of LD between CNVs and SNPs in and around CNVRs is ongoing. This study was supported by the FP7 framework project QUANTOMICS contract n. 222664-2.