Genome wide association study of 40 clinical measurements in eight dog breeds

Abstract

The domestic dog represents an ideal model for identifying susceptibility genes, many of which are shared with humans. In this study, we investigated the genetic contribution to individual differences in 40 clinically important measurements by a genome-wide association study (GWAS) in a multinational cohort of 472 healthy dogs from eight breeds. Meta-analysis using the binary effects model after breed-specific GWAS, identified 13 genome-wide significant associations, three of them showed experimental-wide significant associations. We detected a signal at chromosome 13 for the serum concentration of alanine aminotransferase (ALT) in which we detected four breed-specific signals. A large proportion of the variance of ALT (18.1-47.7%) was explained by this locus. Similarly, a single SNP was also responsible for a large proportion of the variance (6.8-78.4%) for other measurements such as fructosamine, stress during physical exam, glucose, and morphometric measurements. The genetic contribution of single variant was much larger than in humans. These findings illustrate the importance of performing meta-analysis after breed-specific GWAS to reveal the genetic contribution to individual differences in clinically important measurements, which would lead to improvement of veterinary medicine. The domestic dogs share many diseases and phenotypes with human 1. Through two bottlenecks resulting from domestication 2 and the frequent use of specific males, each dog breed shows lower heterogeneity for disease. In humans, higher heterogeneity, meaning multiple variants with different levels of effects influencing the same disease, makes it difficult to identify genetic variants associated with disease 3. Therefore, dogs are considered an ideal model animal for identifying genes and genomic loci underlying diseases and phenotypic variation 1,4. Genome-wide association studies (GWAS) using canine single nucleotide polymorphism (SNP) chips have identified susceptibility genes and genomic loci for several complex diseases 5-10 and for many Mendelian disorders 11-16. The diagnosis of a disease often involves taking many clinical measurements including blood and urine analysis. These measurements directly reflect different aspects of the health of each individual. Individual differences in these measurements are known to be heritable in humans, and GWAS have identified genetic loci explaining these individual differences 17. A systematic analysis of 6,046 dogs has identified breed-specific differences in hematological measurements 18 , suggesting also a genetic contribution to the inter-breed variation for these

Identification of Genetic Susceptibility Factors Associated with Canine Gastric Dilatation-Volvulus

“…A value of λ < 1.1 was considered as a condition to define the absence of significant population stratification. Usually, λ around 1.05 is considered an acceptable value, as reported in some previous large canine GWAS [41][42][43]. Manhattan plots for the relevant signals were generated using a custom R script, including the regional LD computed as pairwise R 2 between the peak SNP and all the other variants located in the range ±250,000 bp from the peak SNP.…”

Man’s best friend in life and death: scientific perspectives and challenges of dog brain banking

“…The tremendous amount of data gained through these commercial testing services can also hold great potentials for research purposes if it can be linked with phenotypic data belonging to the tested dogs through genome wide association studies (GWAS) [71]. A vast number of GWAS studies have already proven the potential of dogs in discovering the genetic background of traits and diseases [72][73][74]. However, GWAS studies mainly rely on DNA isolated from non-invasively collected buccal swabs from living dogs, while the variance in gene expression patterns and regulatory mechanisms remain unrevealed.…”

The Genetic Basis of Obesity and Related Metabolic Diseases in Humans and Companion Animals

“…This means complex traits can be mapped in a breed with fewer individuals and fewer markers than in human populations. Dogs are therefore a compelling model for studying human metabolic disease [ 77 , 239 , 240 , 241 ]. Although, trait mapping in such inbred species means mapping to much larger loci than in humans [ 242 , 243 ], making causative variant identification within an associated locus more difficult.…”