The Use of Population-Scale Sequencing to Identify CNVs Impacting Productive Traits in Different Cattle Breeds

Individualized copy number variation (CNV) maps have highlighted the need for population surveys of cattle to detect rare and common variants. While SNP and comparative genomic hybridization (CGH) arrays have provided preliminary data, next-generation sequence (NGS) data analysis offers an increased resolution and sensitivity for CNV detection. In this study, we analyzed NGS sequence data derived from 67 taurine (consisting of the Angus, Holstein, Jersey, Limousin and Romagnola breeds) and 20 indicine cattle (consisting of the Brahman, Gir and Nelore breeds). Individual genome sequence coverage ranged from 4X to 30X , with an average coverage of 11.8X across animals. To identify CNVs, we used a customized read-depth CNV calling algorithm that utilizes population-scale data to derive a smoothing coefficient for lowess normalization. We identified 2,947 unique CNV regions that account for approximately 4% (117 Mbp) of the cattle genome. Several dairy- and beef-breed specific variants were found, including a duplication of the cell growth-related RICTOR gene in dairy breeds and a duplication of the Patatin-like phospholipase 3 (PNPLA3) gene in beef breeds. Identification of low frequency and breed-specific CNVs within cattle will enable a better understanding of functional variants influenced by domestication and selection for performance traits.