Title


mrrobust -- commands for two-sample Mendelian randomization (MR) analyses.

https://remlapmot.github.io/mrrobust/

Commands

mr Primary command syntax, i.e. use mr egger ... syntax instead of mregger ...

mrdeps Install dependencies for the package

mregger MR-Egger and inverse-variance weighted (IVW) estimators

mreggersimex Simulation extrapolation algorithm for the MR-Egger model

mreggerplot Scatter plot showing instrument specific estimates with IVW, MR-Egger, or median fitted line and confidence interval

mrmedian Unweighted, weighted, and penalized weighted median estimators for summary level data

mrmedianobs Unweighted, weighted, and penalized weighted median estimators for individual level data

mrmodal Modal estimator for summary level data

mrmodalplot Plot of density used in the modal estimator

mrratio Ratio (Wald) estimator for summary level data for a single genotype

mrivests Generate ratio (Wald) estimates for summary level data in dataset

mrforest Forest plot of genotype specific and model (IVW, MR-Egger, Median, Modal) IV estimates

mrfunnel Funnel plot of the genotype specific IV estimates

mrmvivw Multivariable inverse variance weighted estimator (also mvivw and mvmr)

mrmvegger Multivariable MR-Egger regression

mrleaveoneout Leave one out analysis

Description

mrrobust is a suite of programs implementing recently developed estimators which are robust to certain proportions of invalid instrumental variables.

Most of the commands are designed to use summary level data as provided by repositories such as MR-Base.

The estimators were developed in the context of MR studies but could be used for other applications of instrumental variables.

There is a website showing the examples from the helpfiles here: https://remlapmot.github.io/mrrobust/

Author

INCLUDE help mrrobust-author