Colocalization Analysis
Colocalization tests whether two traits share the same causal variant at a given genomic region — critical for linking disease GWAS signals to molecular phenotypes like gene expression or protein levels.
Our pipeline combines the coloc Bayesian framework with SuSiE-based fine-mapping, enabling multi-signal colocalization and substantially improving precision over single-signal methods.
- Standard coloc (H0–H4 posterior probabilities) for single-signal regions
- SuSiE-coloc for multi-signal colocalization at complex loci
- Integration with eQTL catalogues (GTEx, eQTLGen, CAGE)
- pQTL integration (SomaScan, Olink, UKB PPP)
- Prioritisation of high-confidence shared variants
Colocalization Outcomes
01.
H4 > 0.8
Strong evidence for shared causal variant — proceed to mechanism mapping
02.
H3 > 0.5
Distinct causal variants — independent associations, rule out shared signal
03.
H0 / H1 / H2
Insufficient or single-trait signal — flag for additional data or larger GWAS
04.
SuSiE Credible Sets
Multi-signal fine-mapping for regions with multiple independent effects
