Exploring the feasibility of using copy number variants as genetic markers through large-scale whole genome sequencing experiments

Abstract Text: Copy number variants (CNV) are large scale duplications or deletions of genomic sequence that are caused by a diverse set of molecular phenomena that are distinct from single nucleotide polymorphism (SNP) formation. Due to their different mechanisms of formation, CNVs are often difficult to track using SNP-based linkage disequilibrium inference. This can result in decreased reliabilities of prediction for CNV causal mutations tracked by SNP genotyping arrays. To test if CNVs can serve as suitable genetic markers, we sequenced 75 individual bulls from eight different breeds and two subspecies of cattle (Bos taurus taurus: Angus, Holstein, Jersey, Limousin, Romagnola; Bos taurus indicus: Brahman, Gir, Nelore) to 11X coverage. We identified 1,853 non-redundant CNV regions (CNVR) that comprise ~3.1% (87.5 Megabases) of the cattle genome, which represents an increase over previous cattle genome variability estimates (~2%). With the discrete genome copy number values identified in our analysis, we selected the top 1% (n = 80) of CNV sites found to be variable among the sequenced breeds by a modified F statistical measure to perform population structure analyses. We were able to distinctly separate breeds of cattle based on genomic copy number, suggesting that CNVs may have utility as genetic markers. Further analysis revealed that 77.5% (62/80) of our selected CNV windows could reliably be assessed for variability and that 54 of these loci were, in turn, located near tandem duplications. CNV genotyping remains a difficult endeavor and suffers from several obstacles related to their detection and mechanisms of formation; however, these initial results suggest that our current methods can be refined and may provide suitable utility for genomic evaluation in the future. 

Keywords: sequence data, genetic markers, genotyping