LDPAC

LDPAC identifies and quantifies potential spurious associations in genetic association studies by leveraging linkage disequilibrium and likelihood-ratio-based post-association analysis.

Key Features:

• Systematic framework: Implements a formal likelihood-ratio approach to assess the probability that a statistically significant marker reflects a spurious association.
• Utilization of linkage disequilibrium: Evaluates the expectation that true effects produce significant P-values among neighboring markers in LD and flags deviations from this pattern.
• Post-association analysis using P-values and LD: Operates after association testing to compare P-values across neighboring markers informed by LD structure.
• High detection rate: Simulations reported an 84% detection rate for spurious associations.
• Rescue of candidate associations: Identifies potentially interesting candidate associations among markers excluded by initial QC filters.

Scientific Applications:

• Post-association quality control in large-scale genetic studies: Refines association results to reduce false positives in high-throughput GWAS datasets.
• WTCCC type 1 diabetes locus discovery: Applied to Wellcome Trust Case Control Consortium data to identify an additional candidate association for type 1 diabetes among QC-excluded markers that was later confirmed by meta-analysis.

Methodology:

Operates post-association by using P-values and linkage disequilibrium across neighboring markers to compute likelihood ratios that quantify evidence for spurious associations.