Using SNP Weights Derived From Gene Expression Modules to Improve GWAS Power for Feed Efficiency in Pigs

Area of Science:

• Genomics and Bioinformatics
• Animal Genetics and Breeding
• Quantitative Genetics

Background:

• The "large p, small n" problem challenges genome-wide association studies (GWAS), where the number of genetic markers (p) exceeds the number of individuals (n).
• Leveraging prior biological information, such as gene expression patterns, can improve the power and accuracy of GWAS by prioritizing relevant genomic regions and markers.
• Feed efficiency in livestock is a complex trait influenced by numerous genetic factors, making its dissection via traditional GWAS difficult.

Purpose of the Study:

• To develop and validate a novel GWAS methodology that integrates multi-tissue RNA-Seq gene expression data as prior information for SNP weighting and selection.
• To enhance the statistical power of GWAS for identifying significant genetic associations related to feed intake in pigs.
• To compare the efficacy of the proposed weighted GWAS approach against standard unweighted GWAS for detecting biologically relevant SNP markers.

Main Methods:

• RNA-Sequencing (RNA-Seq) data from hypothalamus, duodenum, ileum, and jejunum tissues of pigs with divergent feed efficiency phenotypes were analyzed.
• Constrained tensor decomposition was employed for three-way gene x individual x tissue clustering, yielding 10 gene expression modules.
• Gene module loading values were used to assign weights to 49,691 SNP markers, enabling SNP selection and weighted hypothesis testing in GWAS for feed intake.

Main Results:

• The weighted GWAS identified 36 unique significant SNP associations across 10 gene modules, whereas a standard unweighted GWAS identified only 2 significant SNP.
• A significantly higher proportion (80.6%) of SNP identified through weighted GWAS were located within known quantitative trait loci (QTL) for swine feed efficiency traits compared to unweighted GWAS.
• Specifically, 9 significant SNP from the weighted analysis were located within feed intake QTL, underscoring the biological relevance of the findings.

Conclusions:

• Gene expression data from multiple tissues can effectively serve as prior information to improve SNP prioritization and increase the power of GWAS.
• The proposed weighted GWAS approach is a promising strategy for dissecting the genetic architecture of complex traits, particularly when heritability is driven by many small-effect SNPs.
• This methodology offers a valuable tool for identifying significant genetic variants associated with economically important traits in livestock, facilitating marker-assisted selection and genetic improvement.