BIGKnock

BIGKnock performs gene-based association testing on biobank-scale genotype-phenotype data to prioritize causal genes using knockoff-based conditional genome-wide tests integrated with long-range chromatin interaction data.

Key Features:

• Scalability: Handles biobank-scale datasets and has been applied to the UK Biobank with 405,296 participants.
• Integration of Chromatin Interaction Data: Incorporates long-range chromatin interaction data to link regulatory variants to genes.
• Conditional Genome-Wide Testing via Knockoffs: Implements knockoff statistics for conditional genome-wide testing to distinguish causal signals from correlated proxies.
• Prioritization of Causal Genes: Prioritizes potential causal genes over proxy associations at associated loci.
• Efficiency in Identifying Significant Genes: Produces smaller sets of significant genes with a higher probability of containing causal genes compared to conventional gene-based tests.

Scientific Applications:

• Biobank-scale GWAS: Performs gene-based association testing in large biobank genome-wide association studies.
• Multi-trait Analysis: Analyzes multiple binary and quantitative traits in genotype-phenotype datasets.
• Causal Gene Mapping: Refines locus interpretation by prioritizing causal genes and distinguishing proxy associations within associated regions.

Methodology:

Integrates long-range chromatin interaction data, performs conditional genome-wide testing using knockoff statistics, and prioritizes causal genes over proxy associations.