Skip to content

Repository files navigation

MendelianMR

ci codeql license: MIT python: 3.10+

Browser-based Mendelian Randomization analysis tool.

Methods

  • IVW (fixed + random effects) — primary estimate
  • MR-Egger — directional pleiotropy detection via intercept test
  • Weighted Median — robust when ≥50% weight from valid instruments
  • Simple Median — robust when ≥50% instruments valid
  • MR-PRESSO — outlier detection with global test + Bonferroni correction

Diagnostics

  • Cochran's Q, I² heterogeneity
  • F-statistics for instrument strength
  • Leave-one-out sensitivity analysis

Plots

  • Scatter (SNP-exposure vs SNP-outcome with method lines)
  • Forest (individual Wald ratios + summary diamonds)
  • Funnel (asymmetry detection)
  • Leave-one-out

Example Datasets

  • LDL → CHD (28 SNPs)
  • BMI → T2DM (14 SNPs)
  • SBP → Stroke (8 SNPs)
  • CRP → CHD (5 SNPs, pleiotropy example)

Live Demo

https://mahmood726-cyber.github.io/MendelianMR/

Author

Mahmood Ahmad, Tahir Heart Institute

About

Browser-based Mendelian Randomization analysis tool.

Topics

Resources

Code of conduct

Contributing

Security policy

Stars

Watchers

Forks

Releases

Packages

Contributors

Languages