Genotype imputation in the domestic dog

Area of Science:

• Canine Genetics
• Genomic Data Analysis
• Bioinformatics

Background:

• SNP array genotyping is common in dogs but provides limited genomic coverage.
• Whole genome sequencing offers dense genotype data but is costly.
• Genotype imputation can bridge this gap by predicting dense genotypes from sparse array data.

Purpose of the Study:

• To develop and evaluate genotype imputation methods for dogs.
• To determine optimal reference panel composition for accurate imputation.
• To compare the performance of different imputation software.

Main Methods:

• Developed a reference panel of 4,885,283 SNPs from 83 dogs across 15 breeds.
• Imputed genotypes for 268 dogs using 84,193 SNP array data.
• Performed breed clustering, evaluated reference panel combinations, and compared Beagle and IMPUTE2 software.

Main Results:

• Breed clustering was well-preserved after imputation.
• Multi-breed reference panels yielded higher genotype concordance (92.4%) than breed-specific (87.0%) or non-overlapping panels (74.9%).
• IMPUTE2 showed slightly higher concordance (94.1%) than Beagle (92.4%) with multi-breed panels.

Conclusions:

• Genotype imputation from SNP array data to whole genome level is feasible and accurate in dogs.
• Optimal imputation requires appropriate breed overlap between target and reference panels.
• This method enhances canine genetic analyses by supplementing array-based genotyping data.