gwas

GWAS Module

Genome-Wide Association Studies analysis, fine-mapping, and variant-expression integration.

Architecture

graph TD
    subgraph "GWAS Module"
        A[analysis/] --> |association.py| AS[Association Testing]
        A --> |mixed_model.py| LM[EMMA Mixed Models]
        A --> |quality.py| QC[VCF Parsing & QC]
        A --> |structure.py| ST[PCA & Kinship]

        F[finemapping/] --> |credible_sets.py| CS[Credible Sets & SuSiE]
        F --> |colocalization.py| CO[Colocalization]
        F --> |eqtl.py| EQ[eQTL Analysis]

        D[data/] --> |download.py| DL[Reference Genome & SRA Downloads]
        D --> |genome.py| GN[Chromosome Mapping & GFF3]
        D --> |metadata.py| MD[Sample Metadata]

        H[heritability/] --> |estimation.py| HE[LDSC, GREML, HE Regression]

        V[visualization/] --> VIS[Manhattan, QQ, Interactive Plots]

        W[workflow/] --> WF[Pipeline Orchestration]
    end

Loading

Key Capabilities

Fine-Mapping & Colocalization

from metainformant.gwas.finemapping.colocalization import (
    eqtl_coloc,        # GWAS-eQTL colocalization
    multi_trait_coloc, # Multi-trait analysis
    compute_clpp,      # CLPP method
    regional_coloc,    # Regional analysis
)

# Test if GWAS signal colocalizes with expression QTL
result = eqtl_coloc(
    gwas_z=[1.2, 2.5, 3.1],  # GWAS Z-scores per variant
    eqtl_z=[1.1, 2.3, 2.9],  # eQTL Z-scores
    gene_id="LOC12345"
)
# Returns: PP_H4 (shared causal), interpretation, credible sets

Association Testing

Function	Purpose
run_gwas()	Full GWAS pipeline
logistic_regression()	Case-control associations
linear_regression()	Quantitative trait analysis
lmm_association()	Mixed models for population structure

Data I/O

Function	Purpose
read_vcf()	Parse VCF variant files
read_plink()	Load PLINK bed/bim/fam
write_sumstats()	Export summary statistics

Submodules

Module	Purpose
analysis/	Association testing
data/	VCF, PLINK I/O
finemapping/	Fine-mapping, colocalization
heritability/	Heritability estimation
visualization/	Manhattan, QQ, LocusZoom
workflow/	Pipeline orchestration

Integration with Expression Data

The GWAS module integrates with RNA-seq via the multiomics module:

from metainformant.multiomics.analysis import integration
from metainformant.gwas.finemapping.colocalization import eqtl_coloc

# Convert GWAS and expression data to common format
dna_data = integration.from_dna_variants(gwas_sumstats)
rna_data = integration.from_rna_expression(expression_matrix)

# Run integrated analysis
integrated = integration.integrate_omics_data(
    dna_data=dna_data,
    rna_data=rna_data,
)

Related

• eQTL Integration Pipeline - Maps DNA variants against RNA expression
• metainformant.multiomics - Multi-omic integration
• metainformant.rna - RNA-seq analysis
• config/gwas/ - Configuration files