Bridging GWAS to genes: an integrative multi-omics approach using cattle data.

Area of Science:

• Animal Genomics
• Quantitative Genetics
• Bioinformatics

Background:

• Genome-wide association studies (GWAS) identify genetic loci for complex traits but struggle to pinpoint causal variants and target genes.
• Multi-omics data integration offers a powerful strategy to overcome these challenges.

Purpose of the Study:

• To identify causal genes for milk lactose percentage (LP) in dairy cattle using a multi-breed dataset and multi-omics approach.
• To leverage genomic and transcriptomic data to link genetic variants to gene expression and phenotypic traits.

Main Methods:

• Utilized a large multi-breed dataset (>81,000 cows) with milk LP phenotypes and imputed sequence genotypes.
• Applied BayesR for SNP effect estimation and predicted local genomic breeding values (GEBVs).
• Employed genetic score omics regression (GSOR) and a window-based colocalization test with GWAS summary statistics.

Main Results:

• Identified 711 significant genes (FDR ≤ 0.1) associated with local GEBVs in mammary tissue using GSOR.
• Found 30 significant colocalization windows between GWAS signals and GSOR-identified genes, implicating 34 candidate genes.
• Highlighted 20 genes enriched in 'transmembrane transport' GO terms, relevant to lactose production physiology.

Conclusions:

• The 20 identified genes are strong candidates for causal genes underlying milk lactose percentage, supported by mammary expression, GEBV association, GWAS colocalization, and functional enrichment.
• Demonstrated the effectiveness of integrating GWAS, gene expression, and functional data for causal gene discovery in complex traits.