P0632
Genome-Wide Association Study of a Novel Canine Diversity Panel to Identify QTLs for Morphological Traits in the Domestic Dog

Date: Monday, January 14, 2013
Room: Grand Exhibit Hall
Jessica J Hayward , Biomedical Sciences, Cornell University, Ithaca, NY
Marta Castelhano , Clinical Sciences, Cornell University, Ithaca, NY
Kyle Oliveira , Biomedical Sciences, Cornell University, Ithaca, NY
Elizabeth Corey , Clinical Sciences, Cornell University, Ithaca, NY
Sara E Kalla , Clinical Sciences, Cornell University, Ithaca, NY
Rory Todhunter , Cornell University, Ithaca, NY
Nathan B. Sutter , Clinical Sciences, Cornell University, Ithaca, NY
Adam R. Boyko , Department of Biomedical Sciences Cornell University, Ithaca, NY
The domestic dog (Canis lupus familiaris) displays an extraordinary level of phenotypic diversity, largely due to extensive artificial selection for specific desirable traits. Breed mapping approaches have successfully identified numerous quantitative trait loci (QTLs) underlying breed-stereotyped traits such as fur length, and body size. For most of these phenotypes, a few QTLs of large effect account for the majority of variation among dog breeds.

To further our understanding of the genetic basis of these traits, we genotyped a diverse panel of 2000 dogs representing over 100 breeds at more than 170,000 SNP markers. Whole genome association studies (GWAS) using EMMAX revealed additional loci controlling fur phenotypes as well as associations with previously unmapped traits such as interdigital webbing.

In contrast to previous breed-mapping studies, our panel also contained individual measurements for most dogs. Although pooled GWAS across breeds using individual versus breed-average measures yielded similar results, GWAS within breeds enabled identification of additional loci affecting traits within certain breeds. Analysis of individual body weights across 1260 dogs showed that loci linked to metabolic genes play a substantial role in size variation within some breeds and suggest a role for dogs as a model for human obesity.

Using individual body measurements, we calculated the degree to which the QTLs of large effect among breeds account for variation within breeds and in free-ranging dogs. Furthermore, we test for non-additive and epistatic effects between loci, providing the most complete understanding to date of the genetic architecture of morphological traits in this model system.