GWAS & Selection signatures

Shared Genetic Architecture of Height Across Species: Leveraging Human GWAS to Enhance Cattle Trait Prediction

---

Abstract

Understanding the genetic architecture of complex traits in cattle is central to improving breeding strategies. The rapid progress of human genomics presents an opportunity to inform livestock improvement by leveraging insights from well-characterised human traits. Human height is a classical polygenic trait and provides a valuable framework for testing whether orthologous genomic regions contribute to stature across species. Here, we evaluated whether genomic regions implicated by human height GWAS-regions that collectively account for nearly all common SNP-based heritability of height in humans-are enriched for additive genetic variance in cattle hip height and capture variance in related traits. We mapped 122,111 human height-associated SNPs to the cattle genome, successfully lifting over to 8,446 loci. Of these, 137 actually had SNP at exactly the same location in imputed whole-genome sequence data from 28,351 Bos taurus indicus heifers and cows. We defined trait-informed regions as ±1 kb windows around mapped loci (MAF > 0.01), producing a union set of 196,166 bovine variants. Using a variance-partition framework, we constructed two genomic relationship matrices (GRMs): one from variants in the trait-informed windows (WIN) and one from the remainder of the genome (REST). We fitted two-kernel mixed models including fixed effects for contemporary group, heterozygosity, and the first four principal components, and analysed hip height, weight, and heifer puberty status (pubertal vs non-pubertal) at ~600 days. Enrichment beyond chance was assessed by comparing WIN variance to estimates from five chromosome-matched random SNP sets of equal size using an inverse-variance-weighted null and a Z-test. The WIN set explained 0.12 (±0.014) of phenotypic variance for hip height (total SNP heritability 0.44) and was strongly enriched relative to the random-set null (Z ≈ 4.75; one-sided p ≈ 1à—10⁻⁶). SNP-level cross-species directionality was mixed, consistent with conserved loci harbouring species-specific segregating (tagging) variants rather than identical causal alleles. Mapped regions included biologically plausible growth genes (e.g., HMGA1, IGF2R, MTSS1). For weight, WIN explained 0.0425 (±0.0115) of phenotypic variance with significant enrichment (Z ≈ 2.42; one-sided p ≈ 7.7à—10⁻³). For heifer puberty, WIN explained 0.0538 (±0.0105) with modest enrichment (Z ≈ 1.98; one-sided p ≈ 2.4à—10⁻²). Overall, human height GWAS loci identify cattle genomic regions enriched for additive variance in stature, supporting partially conserved growth biology across species and motivating trait-informed region selection to aid genetic dissection and genomic prediction in livestock.

Keywords: 2026

How to Cite:

Forutan, M. & Hayes, B., (2026) “Shared Genetic Architecture of Height Across Species: Leveraging Human GWAS to Enhance Cattle Trait Prediction”, World Congress on Genetics Applied to Livestock Production Digital Archive 2026(1): 2286280. doi: https://doi.org/10.31274/wcgalp.23903

Rights: 1