Local Joint Testing Improves Power and Identifies Hidden Heritability in Association Studies

Area of Science:

• Genetics
• Statistical Genetics
• Genomic Association Studies

Background:

• Complex human phenotypes are polygenic, often with multiple causal variants per locus.
• Standard genetic association studies analyzing single nucleotide polymorphisms (SNPs) in isolation explain only a small fraction of heritability.
• Existing multi-SNP methods face challenges in multiple-testing correction, genotyping error sensitivity, and optimization for genetic architectures.

Purpose of the Study:

• To develop and validate a local joint-testing procedure that accounts for linkage disequilibrium and improves the detection of associated genetic loci.
• To address limitations of traditional single-SNP association methods in capturing the full genetic architecture of complex traits.

Main Methods:

• Developed a local joint-testing procedure incorporating multiple-testing correction.
• Leveraged the phenomenon of 'linkage masking' where linkage disequilibrium can obscure SNP signals.
• Applied the method to Wellcome Trust Case Control Consortium (WTCCC) data and a cis-expression quantitative trait loci (eQTL) study (gEUVADIS).

Main Results:

• Identified 22 associated loci in WTCCC data, exceeding the marginal approach by 5 loci.
• These newly discovered loci significantly increased the heritability explained by genome-wide significant associations.
• Increased the number of genes with significant cis-eQTLs by 10.7% in the gEUVADIS dataset compared to marginal analyses.

Conclusions:

• Local joint testing offers a more powerful approach for genetic association studies, particularly for complex polygenic traits.
• The developed method effectively overcomes 'linkage masking' and enhances the discovery of disease-associated loci and regulatory variants.
• The Jester software package provides an accessible implementation of this advanced statistical framework.