Bio Causal Genomics Mendelian Randomization

Version Compatibility

Reference examples tested with: TwoSampleMR 0.6.0+, MendelianRandomization 0.10+, MR-PRESSO 1.0+, cause 1.2+, MVMR 0.4+, ieugwasr 1.0+, MRlap 0.0.3.2+, coloc 5.2+, mrclust 0.1+, lhcMR 0.0.1+, R 4.4+. Both TwoSampleMR 0.6.0 and ieugwasr 1.0 are the JWT-transition versions; older versions still expect deprecated OAuth.

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

• R: packageVersion('<pkg>') then ?function_name to verify parameters
• CLI (plink, GCTA-GSMR): <tool> --version then <tool> --help

If code throws an error referencing a function that has moved (e.g. ieugwasr::ld_clump vs TwoSampleMR::clump_data) or an OAuth token failure, introspect the installed API and adapt the example rather than retrying.

Mendelian Randomization

"Test whether trait X causally affects trait Y from GWAS summary statistics" -> Use genetic variants as instrumental variables (IVs) that satisfy three assumptions (relevance, independence, exclusion restriction) to estimate beta_causal = beta_outcome / beta_exposure under the IV framework (Davey Smith & Ebrahim 2003 IJE 32:1; Burgess & Thompson 2021 Chapman & Hall/CRC, 2nd ed.). Tool choice is a decision about the regime (one-sample vs two-sample, sparse vs polygenic, drug-target vs polygenic exposure) and the pleiotropy model (balanced, directional InSIDE, correlated horizontal). Wrong tool inflates Type-I error or attenuates true effects in a direction predictable from the bias structure.

• R: TwoSampleMR::mr() orchestrates IVW + Egger + weighted median + weighted mode in one call
• R: MendelianRandomization::mr_ivw / mr_egger / mr_median / mr_mbe / mr_conmix per-method API (S4 objects; MR-RAPS is NOT in this package -- use TwoSampleMR::mr_raps() which wraps the GitHub mr.raps)
• R: MRPRESSO::mr_presso() global / outlier / distortion tests
• R: cause::cause() correlated horizontal pleiotropy mixture
• R: MVMR::strength_mvmr() + MVMR::ivw_mvmr() multivariable conditional-F + IVW

Statistical Model Taxonomy

Methodology evolves; benchmark consensus shifts every 2-3 years. Verify against the current Slob & Burgess 2020 Genet Epidemiol, Burgess 2023 Wellcome Open Res "Guidelines for performing Mendelian randomization" (v3+), and STROBE-MR 2021 reporting standards before locking a method as primary.

Decision Tree by Experimental Scenario

One-Sample vs Two-Sample Bias Direction

Operational rule: Whenever both GWAS came from UK Biobank (or any single biobank), treat the analysis as one-sample-equivalent and prefer MR-RAPS as primary. Treating it as "two-sample because separate GWAS files" is a common error and produces overestimates.

MRlap: unified correction for sample overlap + winner's curse + weak instruments

MRlap (Mounier & Kutalik 2023 Genet Epidemiol 47:314) jointly corrects three biases that previously required three separate tools: sample overlap, winner's curse, and weak-instrument bias. It builds on an LDSC scaffold (cross-trait LD-score regression intercept estimates the overlap-induced covariance) and reweights the IVW estimate against the analytical bias-correction formula.

remotes::install_github('n-mounier/MRlap')   # never on CRAN; bioconductor unsuitable
library(MRlap)

fit <- MRlap(
    exposure = gwas_X_df, exposure_name = 'BMI',
    outcome = gwas_Y_df, outcome_name = 'T2D',
    ld = 'eur_w_ld_chr/', hm3 = 'w_hm3.snplist',   # LDSC reference files
    MR_threshold = 5e-8, MR_pruning_dist = 500, MR_pruning_LD = 0.05
)
fit$MRcorrection$corrected_effect       # overlap + winner's curse + weak-IV corrected
fit$MRcorrection$corrected_effect_se
fit$LDSC$h2_exp                          # exposure heritability sanity check
fit$LDSC$int_crosstrait                  # cross-trait LDSC intercept; ~0 means no sample overlap

Decision rule -- prefer MRlap when: (a) any sample overlap is suspected, (b) only sumstats are available (no individual-level data for re-running GWAS on disjoint samples), (c) exposure discovery and outcome were both run inside the same biobank (UKB-on-UKB, FinnGen-on-FinnGen). MRlap returns NA / unstable estimates when h^2 < 0.05; in that regime, fall back to Burgess 2016 overlap-corrected IVW plus MR-RAPS for the weak-IV component.

Drug-Target / cis-MR Framework

cis-MR restricts instruments to the cis-regulatory window of the gene encoding the protein/transcript exposure (Schmidt 2020 Nat Commun 11:3255), relaxing the exclusion-restriction assumption because the protein product directly mediates the SNP's effect on the outcome. Operational core: extract cis-pQTL/cis-eQTL within +/-500 kb of the gene; clump at r2 < 0.1 (looser than polygenic MR to retain power within a narrow window); require colocalization PP.H4 >= 0.7; flag protein-altering variants (PAV) which can break SomaScan/Olink aptamer/antibody binding rather than reflect biology.

Full drug-target cis-MR workflow including UKB-PPP / deCODE / Fenland pQTL panels, PAV flagging, Olink vs SomaScan replication (~15-30% cross-platform disagreement), and the operational claim ladder lives in causal-genomics/proteome-mr-drug-target. Use that skill for any drug-target nomination.

Binary outcomes and non-collapsibility

MR with logistic-GWAS sumstats returns per-allele log-OR on the population-averaged (marginal) scale, NOT the conditional log-OR (Burgess 2017 Stat Methods Med Res 26:2333). For rare disease (prevalence < 10%), OR ~= RR ~= HR and the distinction is harmless. For common disease, OR diverges from RR/HR and the MR estimate cannot be back-converted to a conditional effect without strong assumptions; report as "per 1-SD increase in genetically-predicted X, OR for Y = ..." rather than implying an individual-level intervention effect.

Collider bias when conditioning on a collider variable (Coscia 2022 Eur J Epidemiol 37:671 formalizes this for stratified MR): case-only or disease-progression designs condition the sample on disease status, opening a collider path between any cause of disease and any cause of progression. MR within affected subsets without explicit adjustment for selection probability is fragile; weight by inverse probability of selection or restrict claims to the unconditioned population.

Per-Method Failure Modes

IVW under directional pleiotropy

Trigger: Several SNPs affect the outcome through pathways not via the exposure, in a consistent direction.

Mechanism: IVW is a weighted regression through origin; non-zero mean pleiotropy shifts the slope.

Symptom: Egger intercept p < 0.05 with non-zero estimate; IVW differs from weighted median; MR-PRESSO global test p < 0.05.

Fix: Use Egger (if I^2_GX >= 0.9 -- otherwise SIMEX-correct via the simex package applied to the Egger fit, treating se.exposure as measurement error in beta.exposure); cross-check with weighted median, MR-PRESSO, and CAUSE; report IVW only as one of a panel, never alone. The MendelianRandomization::mr_egger() function accepts distribution='normal' and reports the I.sq (I^2_GX) NOME diagnostic but applies no NOME/SIMEX correction to the estimate itself and does NOT expose a SIMEX wrapper.

Weak-instrument bias direction

Trigger: Mean per-instrument F-statistic < 10, or several individual F < 10.

Mechanism: Weak IVs amplify finite-sample correlation between IV-X and IV-Y errors; bias direction depends on overlap regime (see table above).

Symptom: Estimates shift markedly when removing the weakest instruments; one-sample MR estimates much larger than two-sample.

Fix: Compute F per instrument from the EXPOSURE GWAS, not the outcome; exclude F < 10; use MR-RAPS (handles weak IVs by design); for two-sample, also report unweighted IVW (less weak-IV-bias-inflated than weighted in some regimes).

Winner's curse at P~5e-8

Trigger: Discovery GWAS is the source of both instrument selection and effect-size estimates.

Mechanism: SNPs that just cross 5e-8 in discovery have over-estimated effect sizes (regression toward the mean in independent replication); MR uses inflated beta_X, biasing causal estimate.

Symptom: MR effect shrinks substantially when using effect sizes from an independent replication GWAS.

Fix: (1) Three-sample design (discovery / replication-for-instrument-effect / outcome) where feasible. (2) When sumstats-only: MRlap (Mounier 2023), MR-SimSS (sample-splitting from sumstats), or RIVW (Ma 2023 Ann Statist 51:211 -- rerandomized IVW) jointly correct winner's curse + weak IVs + overlap. (3) Jiang 2023 IJE 52:1209 empirical magnitude: variant-level inflation ~50-400% near the genome-wide-significance threshold, dropping to <25% when the minimum P <= 1e-13.

NOME violation invalidating Egger

Trigger: Running MR-Egger with I^2_GX < 0.9.

Mechanism: Egger assumes NO Measurement Error in exposure effect sizes (NOME); when violated, Egger slope is attenuated toward null with reciprocal bias on the intercept.

Symptom: mr_pleiotropy_test() Egger estimate disagrees with weighted median in magnitude but agrees in direction; Isq() function returns <0.9.

Fix: Compute Isq(beta_X, se_X) (Bowden 2016 IJE 45:1961); if <0.9, apply SIMEX correction via simex package or report Egger as exploratory only. The MendelianRandomization package's mr_egger() reports the I.sq (I^2_GX) NOME diagnostic but does not itself apply a NOME/SIMEX correction; SIMEX must be run separately.

Steiger filter false flag under unmeasured confounding

Trigger: Applying steiger_filtering() on traits with unmeasured shared confounders (e.g. SES).

Mechanism: Steiger compares variance explained in exposure vs outcome per SNP; an unmeasured confounder upstream of both produces SNPs that explain more variance in the outcome than the exposure, falsely flagging "reverse causation" (Lutz 2022 Genet Epidemiol 46:139).

Symptom: Many SNPs flagged as wrong-direction yet biology and prior MR support forward causation.

Fix: Treat Steiger as a heuristic, not gospel; cross-validate direction with bidirectional MR (forward + reverse with independent instrument sets); for known-confounder-rich domains (psychiatric traits, SES proxies) use LCV or LHC-MR instead, which jointly model confounders.

Palindromic SNP harmonization

Trigger: SNPs with alleles A/T or C/G near MAF 0.5.

Mechanism: Strand orientation is ambiguous for palindromic SNPs when allele frequencies are intermediate; flipping introduces sign errors that look like pleiotropy.

Symptom: harmonise_data() reports many palindromic SNPs dropped; remaining SNPs show heterogeneity from a handful.

Fix: Default action = 2 (infer from allele frequencies) drops MAF~0.5 palindromes; action = 3 drops ALL palindromes (most conservative); never use action = 1 (assumes forward strand) unless both GWAS are guaranteed to use the same strand convention. Document choice in methods.

Quantitative Thresholds

TwoSampleMR Standard Workflow

Goal: Produce a defensible primary IVW estimate plus a full sensitivity battery from two-sample summary statistics.

Approach: Extract genome-wide significant instruments -> clump (local plink preferred) -> extract outcome -> harmonise -> mr -> pleiotropy + heterogeneity + leave-one-out -> Steiger -> MR-PRESSO -> report.

library(TwoSampleMR)
library(ieugwasr)

exposure_raw <- 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'
)

exposure_sig <- subset(exposure_raw, pval.exposure < 5e-08)  # genome-wide significance

# F-statistic computed from EXPOSURE (Burgess 2011); ratio of squared effect to its variance
exposure_sig$f_stat <- (exposure_sig$beta.exposure / exposure_sig$se.exposure)^2
exposure_sig <- subset(exposure_sig, f_stat >= 10)  # Staiger-Stock 1997 weak-IV heuristic

clumped <- ld_clump(
    data.frame(rsid = exposure_sig$SNP, pval = exposure_sig$pval.exposure),
    clump_r2 = 0.001, clump_kb = 10000,  # polygenic MR convention
    plink_bin = genetics.binaRies::get_plink_binary(),
    bfile = '1kg_EUR/EUR'
)
exposure_dat <- subset(exposure_sig, SNP %in% clumped$rsid)

outcome_dat <- read_outcome_data(
    filename = 'outcome_gwas.tsv', snps = exposure_dat$SNP, 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'
)

dat <- harmonise_data(exposure_dat, outcome_dat, action = 2)  # infer from EAF; drops MAF~0.5 palindromes

primary <- mr(dat, method_list = c('mr_ivw', 'mr_egger_regression',
                                    'mr_weighted_median', 'mr_weighted_mode'))

heterogeneity <- mr_heterogeneity(dat)         # Cochran Q
pleiotropy <- mr_pleiotropy_test(dat)          # Egger intercept
loo <- mr_leaveoneout(dat)                     # influential-SNP check
steiger <- directionality_test(dat)            # variance-explained direction

MR-PRESSO Outlier Detection

Goal: Detect horizontal-pleiotropy outliers, remove them, and test whether the corrected estimate differs from the uncorrected one (distortion test).

Approach: Three-step framework: global test (presence of pleiotropy), outlier test (per-SNP), distortion test (effect change after outlier removal).

library(MRPRESSO)

presso <- mr_presso(
    BetaOutcome = 'beta.outcome', BetaExposure = 'beta.exposure',
    SdOutcome = 'se.outcome', SdExposure = 'se.exposure',
    OUTLIERtest = TRUE, DISTORTIONtest = TRUE,
    data = dat, NbDistribution = 10000,  # >= 10000 for publication-grade p-value precision
    SignifThreshold = 0.05
)

print(presso$`MR-PRESSO results`$`Global Test`)         # any pleiotropy
print(presso$`MR-PRESSO results`$`Distortion Test`)     # change after outlier removal
outlier_snps <- which(presso$`MR-PRESSO results`$`Outlier Test`$Pvalue < 0.05 / nrow(dat))

CAUSE for Correlated Horizontal Pleiotropy

CAUSE (Morrison 2020 Nat Genet 52:740) fits a shared-factor mixture to genome-wide sumstats and compares causal vs sharing-only models by delta-ELPD. Workflow: gwas_merge() -> sample ~1M variants for est_cause_params() -> filter sig SNPs (P < 1e-3) and optionally LD-prune -> cause(X, variants, param_ests). Needs >= 100 sig SNPs. Full annotated example and ELPD interpretation in causal-genomics/pleiotropy-detection.

MVMR with Conditional F

Goal: Estimate the causal effect of exposure X1 on Y, adjusting for measured pleiotropy via X2.

Approach: Format exposures + outcome into MVMR object; compute conditional F per exposure (>10 required); run multivariable IVW; report Q_A heterogeneity.

library(MVMR)

mvmr_dat <- format_mvmr(
    BXGs = cbind(dat$beta.x1, dat$beta.x2),
    BYG = dat$beta.y,
    seBXGs = cbind(dat$se.x1, dat$se.x2),
    seBYG = dat$se.y,
    RSID = dat$SNP
)

condF <- strength_mvmr(r_input = mvmr_dat, gencov = 0)  # per-exposure conditional F
# condF must be > 10 for EACH exposure (Sanderson 2019); total F is misleading

mv_ivw <- ivw_mvmr(r_input = mvmr_dat)
mv_qa <- pleiotropy_mvmr(r_input = mvmr_dat, gencov = 0)  # Q_A heterogeneity test

gencov = 0 is valid ONLY if the exposure GWAS samples don't overlap; for overlapping exposures use the bivariate LDSC intercept matrix as gencov. If any conditional F < 10, the IVW point estimate is weak-IV-biased; switch to the Q-minimization estimator: qhet_mvmr(r_input, pcor, CI = TRUE, iterations = 1000) (Sanderson 2021 Stat Med 40:5434), which minimizes Q-statistic heterogeneity rather than weighting by inverse variance and is robust to weak conditional instruments.

Bidirectional and Steiger

exposure_rev <- format_data(outcome_raw, type = 'exposure')  # treat former outcome as exposure
outcome_rev <- format_data(exposure_raw, type = 'outcome')
dat_rev <- harmonise_data(exposure_rev, outcome_rev, action = 2)
results_rev <- mr(dat_rev, method_list = 'mr_ivw')

dat_filt <- steiger_filtering(dat)  # per-SNP; flags SNPs where variance(Y) > variance(X)
dir_test <- directionality_test(dat) # global; correct_causal_direction == TRUE if forward

Report direction operationally: forward p < 5e-8 with reverse p > 0.05; directionality_test() correct_causal_direction == TRUE with Steiger p < 0.05; point estimate in reverse direction has |effect| substantially smaller than forward (reflecting reverse-instrument-strength asymmetry). Example sentence: "Forward MR showed BMI -> T2D (IVW beta = 0.85, p = 2e-15); reverse MR was null (IVW beta = 0.02, p = 0.61); Steiger directionality test favored the forward direction (p = 3e-7)."

Reconciliation: When Methods Disagree

Operational rule for publication: Primary IVW + concordant Egger (or weighted median if NOME violated) + non-significant MR-PRESSO global test + Steiger correct direction = publication-ready evidence. CAUSE concordance is required when the exposure is polygenic and the prior on CHP is high (e.g. BMI -> outcome, lipids -> outcome). Single-method "significant IVW" claims should be downgraded to exploratory.

Cohort Gotchas

• UKB GWAS commonly include non-EUR participants; pan-UKB EUR/AFR/EAS/SAS subsets are separate releases. Mixing ancestries inflates instrument strength via population stratification rather than biology.
• FinnGen DF12 (2024) cohort: Finnish founder effects produce narrower LD blocks and higher winner's curse magnitude than UKB at matched sample size.
• MVP / GBMI / AoU: multi-ancestry meta-analyses; stratify by ancestry before MR or use ancestry-specific subsets.
• UKB-on-UKB MR creates one-sample-equivalent bias regardless of "different GWAS file" appearance; use MRlap or move the outcome to an external cohort (FinnGen, BBJ, MVP).

Anticipated Reviewer Pushback

Common Errors

Tool Installation Notes

# CRAN-stable
install.packages(c('remotes', 'MendelianRandomization', 'MVMR', 'coloc'))

# GitHub-only or recently archived
remotes::install_github('MRCIEU/TwoSampleMR')          # primary orchestrator
remotes::install_github('MRCIEU/ieugwasr')             # OpenGWAS client + local clumping
remotes::install_github('rondolab/MR-PRESSO')          # never on CRAN
remotes::install_github('qingyuanzhao/mr.raps')        # CRAN-archived 2025-03-01
remotes::install_github('jean997/cause')               # depends on mixsqp; suggests Rfast
remotes::install_github('cnfoley/mrclust')             # heterogeneity clusters
remotes::install_github('LizaDarrous/lhcMR')           # bidirectional + heritable confounder
remotes::install_github('HDTian/DRMR')                 # doubly-ranked stratification
remotes::install_github('n-mounier/MRlap')             # joint overlap + winner's-curse + weak-IV correction

TwoSampleMR::mr_raps() is a thin wrapper that calls mr.raps::mr.raps() under the hood; the GitHub mr.raps install above is therefore required. The MendelianRandomization package does NOT export mr_raps() (verify with ls('package:MendelianRandomization')); only TwoSampleMR offers a MR-RAPS entry point. For local clumping, install plink2 and download a 1KG EUR (or matched-ancestry) reference bfile.

STROBE-MR Reporting

STROBE-MR (Skrivankova 2021 JAMA 326:1614; BMJ 375:n2233): 20-item checklist required by Eur J Epi / Nat Genet / JAMA / Diabetologia since 2022. See causal-genomics/pleiotropy-detection for the per-item table -- that skill is the consolidated reporting reference for the full MR + sensitivity battery.

References

• Davey Smith G & Ebrahim S 2003 IJE 32:1 (foundational framework)
• Burgess S & Thompson SG, Mendelian Randomization: Methods for Causal Inference Using Genetic Variants, Chapman & Hall/CRC (1st ed. 2015 / 2nd ed. 2021) (MR canonical reference)
• Bowden J et al 2015 IJE 44:512 (MR-Egger)
• Bowden J et al 2016 Genet Epidemiol 40:304 (weighted median)
• Hartwig FP et al 2017 IJE 46:1985 (weighted mode)
• Zhao Q et al 2020 Ann Stat 48:1742 (MR-RAPS)
• Morrison J et al 2020 Nat Genet 52:740 (CAUSE)
• Zhu Z et al 2018 Nat Commun 9:224 (GSMR + HEIDI)
• Verbanck M et al 2018 Nat Genet 50:693 (MR-PRESSO)
• Sanderson E et al 2019 IJE 48:713 (MVMR conditional F)
• Foley CN et al 2021 Bioinformatics 37:531 (MR-Clust)
• Burgess S et al 2020 Nat Commun 11:376 (contamination mixture)
• O'Connor LJ & Price AL 2018 Nat Genet 50:1728 (LCV)
• Darrous L et al 2021 Nat Commun 12:7274 (LHC-MR)
• Tian H et al 2023 PLoS Genet 19:e1010823 (DRMR)
• Burgess S et al 2016 Genet Epidemiol 40:597 (sample-overlap correction)
• Schmidt AF et al 2020 Nat Commun 11:3255 (drug-target / cis-MR framework)
• Lutz SM et al 2022 Genet Epidemiol 46:139 (Steiger filter caveat)
• Skrivankova VW et al 2021 JAMA 326:1614 (STROBE-MR statement)
• Mounier N & Kutalik Z 2023 Genet Epidemiol 47:314 (MRlap joint correction)
• Sanderson E et al 2021 Stat Med 40:5434 (qhet_mvmr Q-minimization estimator)
• Coscia C et al 2022 Eur J Epidemiol 37:671 (collider bias in case-only / progression cohorts)
• Hamilton FW et al 2023 medRxiv 23293658 (NLMR stratum-specific bias critique)
• Jiang T et al 2023 IJE 52:1209 (winner's curse empirical magnitude)
• Ma X et al 2023 Ann Statist 51:211 (RIVW rerandomized IVW)

Related Skills

• causal-genomics/colocalization-analysis - Confirm shared causal variant for cis-MR drug-target work
• causal-genomics/pleiotropy-detection - Deep dive on MR-PRESSO, Egger, contamination mixture diagnostics
• causal-genomics/fine-mapping - Credible-set construction at instrument loci before cis-MR
• causal-genomics/mediation-analysis - Two-step MR and MVMR difference method for X -> M -> Y mediation
• causal-genomics/transcriptome-wide-association - TWAS as MR-adjacent framework for gene-level inference
• causal-genomics/proteome-mr-drug-target - Dedicated cis-pQTL drug-target MR workflow with UKB-PPP/deCODE/Fenland and coloc triangulation
• population-genetics/association-testing - GWAS source for instrument selection
• clinical-databases/clinvar-lookup - Annotate instrument SNPs for downstream interpretation