Five genetic variants explain over 70% of hair coat pheomelanin intensity variation in purebred and mixed breed domestic dogs - Supporting information

In mammals, the pigment molecule pheomelanin confers red and yellow color to hair, and the intensity of this coloration is caused by variation in the amount of pheomelanin. Domestic dogs exhibit a wide range of pheomelanin intensity, ranging from the white coat of the Samoyed to the deep red coat of the Irish Setter. While several genetic variants have been associated with specific coat intensity phenotypes in certain dog breeds, they do not explain the majority of phenotypic variation across breeds. In order to gain further insight into the extent of multigenicity and epistatic interactions underlying coat pheomelanin intensity in dogs, we leveraged a large dataset obtained via a direct-to-consumer canine genetic testing service. This consisted of genome-wide single nucleotide polymorphism (SNP) genotype data and owner-provided photos for 3,057 pheomelanic mixed breed and purebred dogs from 63 breeds and varieties spanning the full range of canine coat pheomelanin intensity. We first performed a genome-wide association study (GWAS) on 2,149 of these dogs to search for additional genetic variants that underlie intensity variation. GWAS identified five loci significantly associated with intensity, of which two (CFA15 29.8 Mb and CFA20 55.8 Mb) replicate previous findings and three (CFA2 74.7 Mb, CFA18 12.9 Mb, CFA21 10.9 Mb) have not previously been reported. In order to assess the combined predictive power of these loci across dog breeds, we used our GWAS data set to fit a linear model, which explained over 70% of variation in coat pheomelanin intensity in an independent validation dataset of 908 dogs. These results introduce three novel pheomelanin intensity loci, and further demonstrate the multigenic nature of coat pheomelanin intensity determination in domestic dogs.

Methods
Cheek cell samples were collected by dog owners with buccal swabs, and DNA was extracted by Illumina, Inc. and genotyped at 221,188 biallelic autosomal and X chromosome markers on the Embark Veterinary custom Illumina CanineHD SNP array. Several filtering steps were performed to remove individuals and markers that did not meet study requirements (these steps are detailed in the linked manuscript). After these steps, the total genotyping rate was 99.9% across 204,571 markers in 3,057 dogs from 63 different breeds and varieties. Dogs were phenotyped from owner-provided photographs (see linked manuscript for details). Genotype and phenotype data were used as input to a series of genome-wide association studies to identify genomic regions associated with pheomelanin intensity variation, which were run using GEMMA v.0.98.