Skip to main content
ARS Home » Pacific West Area » Pullman, Washington » Animal Disease Research » Research » Publications at this Location » Publication #294043
Title: Extent of linkage disequilibrium in large breed dogs: chromosomal and breed variation

Author
item STERN, JOSHUA - North Carolina State University
item White, Stephen
item MEURS, KATHRYN - North Carolina State University

Submitted to: Mammalian Genome
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 8/1/2013
Publication Date: 10/1/2013
Citation: Stern, J., White, S.N., Meurs, K. 2013. Extent of linkage disequilibrium in large breed dogs: chromosomal and breed variation . Mammalian Genome. 24(9-10):409-415

Interpretive Summary: Linkage disequilibrium (LD) is a measure of the tendency of DNA variants on the same chromosome to be inherited together, and knowing the physical extent of LD is critical for discovery of genes and mutations involved in a wide range of inherited traits. This study provided a high resolution analysis of LD in 3 large dog breeds to facilitate further genome-wide association scans (GWAS) for many traits of interest in dogs. DNA from 24 Golden Retriever dogs, 28 Rottweiler dogs and 24 Newfoundland dogs was used for analysis by genome-wide, high density (>170,000) single nucletoide polymorphism (SNP) array. All dogs were healthy and unrelated based upon three-generation pedigree. Extent of LD was significantly shorter for Newfoundland dogs based upon 50% decay of r2 data at a mean of 344 KB compared to Golden Retriever and Rottweiler dogs at 715 KB and 834 KB, respectively (P<0.0001). Notable differences in LD by chromosome were present and not strictly related to respective chromosome lengths. To our knowledge, this is the first report of SNP-based LD for Newfoundland dogs, the first report based on genome-wide SNPs for Rottweilers, and an almost 10-fold improvement in marker density over previous genome-wide studies of LD in Golden Retrievers. The extent of LD is breed-specific and LD further varied by chromosome. This understanding may inform more efficient use of GWAS data for identification of candidate genes and ultimately mutation discovery.

Technical Abstract: Purpose: Understanding extent of linkage disequilibrium (LD) is a crucial component for successful utilization of genome-wide association studies (GWAS). The extent of LD in the dog has been described based upon small marker sets in multiple breeds and studies. Understanding variation in LD on a per chromosome and breed basis by high-density genome-wide coverage allows researchers to better define regions of interest and target candidate genes through a GWAS approach. Procedures: 24 golden retriever dogs, 28 Rottweiler dogs and 24 Newfoundland dogs were genotyped for single nucleotide polymorphism (SNP) data using a high density (>170,000) SNP array. All dogs were healthy and unrelated based upon three-generation pedigree. After filtering, LD was calculated for all autosomes using Haploview. Decay of the squared correlation coefficient (r2) was plotted on a per breed and per chromosome basis as well as in a genome-wide fashion. The difference in extent of LD between breeds was estimated by 50% decay of r2 was reported on a per chromosome and per breed basis. Findings: Extent of LD was significantly shorter for Newfoundland dogs based upon 50% decay of r2 data at a mean of 344 KB compared to golden retriever and Rottweiler dogs at 715 KB and 834 KB respectively (P<0.0001). Notable differences in LD by chromosome were present within each breed and not strictly related to the respective length of the canine autosomes. Conclusions: To our knowledge, this is the first report of SNP-based linkage disequilibrium for Newfoundland dogs, the first report based on genome-wide SNPs for Rottweilers, and an almost 10-fold improvement in marker density over previous genome-wide studies of LD in Golden Retrievers. Although the chosen breeds are similar in phenotype and are over-represented for several congenital diseases, their extent of LD is breed-specific. Additionally LD varied by chromosome. This understanding may inform more efficient use of GWAS data for identification of candidate genes and ultimately mutation discovery.