Multiple-trait QTL mapping and genomic prediction for wool traits in sheep

Area of Science:

• Animal Genetics
• Quantitative Genetics
• Sheep Breeding

Background:

• Genomic selection offers potential for increased profitability in wool production through accurate breeding values derived from single nucleotide polymorphism (SNP) data.
• Key wool traits and those influencing susceptibility to fleece rot and fly strike are crucial for sheep value.
• The study aimed to predict genomic estimated breeding values (GEBV) and compare multi-trait methods for mapping trait-affecting polymorphisms.

Purpose of the Study:

• To predict genomic estimated breeding values (GEBV) for various wool and conformation traits in Merino sheep.
• To compare three multi-trait analysis methods for mapping quantitative trait loci (QTL) with pleiotropic effects.
• To identify genetic polymorphisms influencing economically important wool production and quality traits.

Main Methods:

• Calculated GEBV for 5726 sheep across 22 traits using BayesR and genomic best linear unbiased prediction (GBLUP) with 510,174 SNPs.
• Assessed GEBV accuracies via fivefold cross-validation.
• Devised and compared three approximate multi-trait analyses (including multi-trait GWAS and two BayesR-based methods) to map pleiotropic QTL.

Main Results:

• BayesR and GBLUP yielded similar average GEBV accuracies (~0.22), with BayesR excelling in wool yield and fibre diameter (>0.40).
• Accuracy was generally higher for traits with larger reference populations and heritability.
• The multi-trait analyses identified 206 putative QTL, with 20 common across all methods. BayesR approaches provided more refined QTL mapping than GWAS.

Conclusions:

• The mean accuracy of genomic prediction for wool traits was approximately 0.22.
• 206 putative QTL were identified across the ovine genome using multi-trait analyses.
• Candidate genes (e.g., FGF5, STAT3) associated with hair growth were identified near SNPs with pleiotropic effects, supported by detailed phenotypic data.