All sheeps and sizes: a genetic investigation of mature body size across sheep breeds reveals a polygenic nature

Abstract

Mature body size is genetically correlated with growth rate, an important economic trait in the sheep industry. Mature body size has been studied extensively in humans as well as cattle and other domestic animal populations but not in sheep. Six-hundred and sixteen ewes, across 22 breeds, were measured for 28 linear measurements representing various skeletal parts. PCA from these measures generated principal components 1 and 2 which represented 66 and 7% of the phenotypic variation respectively. Two-hundred and twenty sheep were genotyped on the Illumina Ovine HD beadchip for a GWAS investigating mature body size and linear body measurements. Forty-six (Bonferroni P < 0.05) SNP associations across 14 chromosomes were identified utilizing principal component 1, representing overall body size, revealing mature body size to have fewer loci of large effect than other domestic species such as dogs and horses. Genome-wide associations for individual linear measures identified major quantitative trait loci for withers height and ear length. Withers height was associated (Bonferroni P < 0.05) with 12 SNPs across six chromosomes whereas ear length was associated with a single locus on chromosome 3, containing MSRB3. This analysis identified several loci known to be associated with mature body size in other species such as NCAPG, LCORL, and HMGA2. Mature body size is more polygenic in sheep than other domesticated species, making the development of genomic selection for the trait the most efficient option for maintaining or reducing mature body size in sheep.

Keywords: complex trait; ear length; genome-wide association studies; principal component analysis; withers height.

Figure 1

Factor loadings for individual linear measures within principal components 1 and 2. A blue bar indicates a positive loading value and a red bar indicates a negative loading value. (a) Principal component 1, representing overall body size, shows all positive loading values for linear measures. (b) Principal component 2 generally distinguishes body thickness from length

Figure 2

Boxplot of linear body measure principal component 1 scores sorted by median values for each breed. Only breeds which had five or more ewes were included in this figure. Whiskers indicate 1.5× the interquartile range, whereas open circles indicate sampled data points

Figure 3

Scatterplot of principal components 1 and 2 from linear body measures (n = 607 excluding breeds with fewer than five samples per breed). Principal component 1 is represented on the x‐axis and principal component 2 is represented on the y‐axis with individual animals colored by breed. Percentage values in parentheses represent the percentage of phenotypic variation explained by the principal component

Figure 4

Manhattan plot showing −log₁₀(P‐values) of the association with principal component 1, representing mature body size, in the across‐breed analysis. The horizontal blue line indicates the Bonferroni corrected threshold of 0.05

Figure 5

Manhattan plot showing −log10(P‐values) of the association with withers height in the across‐breed analysis. The horizontal blue line indicates the Bonferroni corrected threshold of 0.05

Figure 6

(a) Manhattan plot showing −log₁₀(P‐values) of the association with ear length in the across‐breed analysis. (b) Manhattan plot showing −log₁₀(P‐value) of the association with ear length which included principal component 1 as a covariate in the model. The horizontal blue line indicates the Bonferroni corrected threshold of 0.05