bio-workflows-causal-genomics-pipeline

Version Compatibility

Reference examples tested with: TwoSampleMR 0.5+, coloc 5.2+, susieR 0.12+

Before using code patterns, verify installed versions match. If versions differ:

• R: packageVersion("<pkg>") then ?function_name to verify parameters

If code throws ImportError, AttributeError, or TypeError, introspect the installed package and adapt the example to match the actual API rather than retrying.

Causal Genomics Pipeline

"Run post-GWAS causal inference from summary statistics" → Orchestrate instrument selection, Mendelian randomization, pleiotropy sensitivity analyses, colocalization, fine-mapping, and mediation analysis to identify causal exposures and mechanisms.

Complete post-GWAS causal inference workflow from summary statistics to causal mechanisms.

Pipeline Overview

GWAS Summary Statistics (exposure + outcome)
    |
    v
[1. Instrument Selection] ----> LD clumping, F-stat filtering
    |
    v
[2. Mendelian Randomization] -> IVW, MR-Egger, Weighted Median
    |
    +--> [3. Sensitivity Analysis] ---> MR-PRESSO, Egger intercept,
    |                                    leave-one-out, Steiger
    |
    v                           (can run in parallel with MR)
[4. Colocalization] ----------> coloc, eCAVIAR, HyPrColoc
    |
    v
[5. Fine-Mapping] ------------> SuSiE, FINEMAP (credible sets)
    |
    v
[6. Mediation Analysis] ------> Network MR, MVMR
    |
    v
Causal evidence summary (triangulation across methods)

Step 1: Instrument Selection

library(TwoSampleMR)

exposure_dat <- read_exposure_data(
    filename = 'exposure_gwas.tsv',
    sep = '\t',
    snp_col = 'SNP',
    beta_col = 'BETA',
    se_col = 'SE',
    effect_allele_col = 'A1',
    other_allele_col = 'A2',
    eaf_col = 'EAF',
    pval_col = 'P'
)

# Genome-wide significance threshold for instrument selection
# p < 5e-8: Standard GWAS threshold; use 5e-6 for underpowered exposures
exposure_dat <- subset(exposure_dat, pval.exposure < 5e-8)

# LD clumping to ensure independent instruments
# clump_r2 = 0.001: Strict independence; 0.01 for more instruments
# clump_kb = 10000: 10 Mb window, standard for MR
exposure_dat <- clump_data(exposure_dat, clump_r2 = 0.001, clump_kb = 10000)