# Genomic prediction of beef quality using GWAS-prioritized markers

## TL;DR

This study shows that using a small set of key genetic markers can accurately predict beef quality traits like tenderness and marbling in cattle.

## Contribution

The study demonstrates that GWAS-prioritized SNP panels can achieve comparable prediction accuracy to full panels for beef quality traits.

## Key findings

- For tenderness (WBSF), small SNP panels (e.g., top 50 SNPs) achieved accuracy similar to full panels due to a major QTL on chromosome 29.
- Marbling prediction required more SNPs, reflecting its polygenic nature.
- GWAS-ranked SNP subsets outperformed random subsets and matched full panel accuracy in several cases.

## Abstract

Tenderness and marbling are key carcass quality traits in beef cattle that strongly influence consumer eating satisfaction and repeat purchasing behavior. Because both traits are measured postmortem, they are difficult to incorporate into routine selection programs. Genomic selection therefore provides a practical strategy to improve these traits. This study evaluated the effectiveness of GWAS-informed SNP preselection for predicting Warner–Bratzler shear force (WBSF) and marbling breeding values using reduced marker densities in Brangus cattle. Using a structured population (N = 1066), we conducted a 10-fold cross-validation with SNP subsets ranked by GWAS significance and compared them to 10 random SNP subsets of the same number and full SNP panel. External validation was performed using 338 animals from an independent source. For WBSF, small panels (eg top 50 SNPs) achieved accuracy comparable to the full panel, driven by a strong QTL on chromosome 29 (CAPN1). In contrast, marbling required broader marker coverage for optimal prediction, consistent with a polygenic trait architecture. Across all subset sizes, GWAS-ranked SNPs outperformed random subsets, and in several cases matched full panel accuracy. External validation confirmed the reliability of these results. These results demonstrate that trait-specific genetic architecture strongly influences the marker density required for reliable genomic prediction and highlight the value of GWAS-informed SNP prioritization for optimizing genomic prediction strategies in crossbred cattle.

Reduced GWAS-informed SNP sets deliver competitive prediction for tenderness and highlight trait-specific differences in marker count requirements.