Genomic Variation in Dogs and Wolves: Discovery, Genotyping, and Breakpoint Analysis of Structural Variants

Understanding canine genome structural variation will clarify important aspects of canine evolution and also shed light on important canine diseases and traits. Using various computational approaches, including read-depth, read-pair, split-read, and local assembly, we have identified structural variants and their putative breakpoints in a diverse collection of over eighty dog and wolf genomes. We detect nearly 20,000 reference deletions with resolved breakpoints, including 2,151 and 6,018 polymorphic LINE and SINE deletions, respectively. We also identify 8,731 deletions that display microhomology at the breakpoints, which suggests origins from homology-based mechanisms (NAHR or NHEJ). In contrast, deletions likely caused by DNA replication-based mechanisms were found in 2,998 instances, as non-reference insertions at the breakpoints were present. In addition, similar methods and subsequent filtration steps found 2,310 tandem duplications, and multiple read-depth based approaches highlighted 8,658 regions of copy-number variation across the canine genomes. Finally, 7,268 contigs of novel sequence absent from the reference genome (over 8 Mb) were assembled from a Great Dane and Grey Wolf, providing the ability to improve the reference and expand the catalog of variation segregating among dogs and wolves. Validation and genotyping of the discovered deletions, tandem duplications, copy number variants, and mobile elements will be accomplished using comparative genomic hybridization arrays (aCGH). The results from this study will highlight the abundance, size, and distribution of structural variants in canine genomes and provide a resource for the imputation of structural and copy-number variation into ongoing studies of canine phenotypes.