One Sample Mendelian Randomization and Instrumental Variable Analyses.
OneSampleMR: Useful functions for one-sample Mendelian randomization and instrumental variable analyses
Useful functions for one sample, individual level data, Mendelian randomization / instrumental variable analyses, including implementations of:
- The Sanderson and Windmeijer (2016) conditional F-statistic for multiple exposure models.
- Various one-sample instrumental variable estimators including the
- Multiplicative structural mean model (Robins, 1989; Hernán and Robins, 2006)
- Two-stage predictor substitution estimators (Terza, 2008)
- Two-stage residual inclusion estimators (Terza, 2008)
OneSampleMR was included in the R Views November 2021 "Top 40" New CRAN Packages here!
Installation
Released version
Install the package from CRAN using
install.packages("OneSampleMR")
To check for updates run
update.packages()
Issue with gmm package versions 1.6-5, 1.6-6, 1.7 on macOS
Using these old versions of the gmm package, a GNU Fortran installation was required even if the binary version of the package was installed.
The fix is to update to the current version of gmm, which is version 1.8.
If you cannot update your version of gmm then
- The current version of GNU Fortran to install is documented here and available here
- Note that the version of Fortran available from Homebrew will probably not work
Development version
Install the development version of the package from source from GitHub using
# install.packages("remotes") # uncomment on first run
remotes::install_github("remlapmot/OneSampleMR")
To check for and install updates to the development version run the previous command again or run
remotes::update_packages()
A binary of the development version of the package (for Windows, macOS, and Ubuntu Jammy Jellyfish) can be installed from r-universe using
install.packages('OneSampleMR', repos = c('https://remlapmot.r-universe.dev', 'https://cloud.r-project.org'))
which can be updated using
update.packages(repos = c('https://remlapmot.r-universe.dev', 'https://cloud.r-project.org'))