LOCATER

LOCATER leverages genealogy-based methods to test tree structures for association with traits, improving association power at loci with allelic heterogeneity by combining signals present on different branches of a locus-specific haplotype tree.

Key Features:

• Tree-based association testing: Tests tree structures represented via discrete clades or relatedness matrices for association with traits.
• Signal combination across branches: Combines signals present on different branches of a locus-specific haplotype tree to boost association power.
• Handling of allelic heterogeneity: Improves power when multiple causal variants are present by testing combinations of common and rare haplotypes based on ancestral relationships.
• Confounder correction: Implements a new randomization approach and a general method for genomic control to eliminate confounder effects that are more pronounced in genealogy-based methods than single marker tests.

Scientific Applications:

• Genome-wide association studies: Applied to a genome sequencing study of 6,795 Finnish individuals with 101 cardiometabolic traits and 18.9 million autosomal variants, identifying 351 significant trait associations at 47 distinct genomic loci.
• Recovery of known quantitative trait loci: Successfully recovers known QTL not found by single marker tests, including LIPG and known allelic heterogeneity at the APOE/C1/C4/C2 gene cluster.
• Noncoding region analysis: Addresses the challenge of improving trait association power in noncoding regions where variant impacts are difficult to predict and functional units are hard to define for variant aggregation.
• Common and rare variant integration: Bridges the gap between common and rare variant analysis by testing haplotype combinations based purely on ancestral relationships without requiring predefined functional units.

Methodology:

LOCATER uses an efficient local ancestry inference engine and a novel statistical method for combining signals on different branches of a locus-specific haplotype tree, with a genome-wide analysis pipeline that includes a randomization approach and genomic control for confounder correction.