Accounting for Population Structure and Phenotypes From Relatives in Association Mapping for Farm Animals: A...

Area of Science:

• Population Genetics
• Genomic Association Studies
• Quantitative Genetics

Background:

• Genome-wide association studies (GWAS) require accounting for population structure to prevent spurious associations.
• Existing GWAS methods are often limited to individuals with both genotypes and phenotypes.
• Single-step GWAS (ssGWAS) offers potential for including phenotypes from non-genotyped relatives but its population structure handling is unexplored.

Purpose of the Study:

• To investigate the performance of single-step GWAS (ssGWAS) in accounting for population structure.
• To compare ssGWAS with established methods like efficient mixed-model association expedited (EMMAX) and genomic best linear unbiased prediction GWAS (GBLUP-GWAS).
• To evaluate differences from single-SNP analysis without population structure correction (SSA-NoCor).

Main Methods:

• Simulated structured populations mimicking fish, beef cattle, and dairy cattle.
• Genomes with 29 chromosomes, each with one quantitative trait nucleotide (QTN).
• Varying numbers of SNPs and genotyped individuals; phenotypes available for a trait with 0.25-0.35 heritability.

Main Results:

• SSA-NoCor identified the most true positives but had a high false negative rate.
• GBLUP-GWAS and EMMAX showed similar true positive rates, slightly lower than ssGWAS.
• No significant differences were found when using deregressed proofs in EMMAX versus daughter phenotypes in ssGWAS for dairy cattle.

Conclusions:

• Single-step GWAS (ssGWAS) effectively accounts for population structure in association analyses.
• ssGWAS is a straightforward method when only a fraction of the population is genotyped.
• ssGWAS is suitable when phenotypes are available on non-genotyped relatives, enhancing genetic studies.