A preprint for the MR-PATH paper is available at https://arxiv.org/abs/2007.06476.
To install the package, run the following command in R.
library(devtools)
install_github("danieliong/MRPATH")
# To automatically update/download dependency packages
install_github("danieliong/MRPATH", upgrade=TRUE)
The main functions in MRPATH are MR_PATH, MRPATH_optimizeInitVals, and MRPATH_selectModel. Each of these functions provide increasing levels of automation in the model-fitting process. In MR_PATH, the user must input data, the desired number of clusters, and initial values. In MRPATH_optimizeInitVals, the user has to input data and the desired number of clusters. In MRPATH_selectModel, the user just has to input data.
Auxiliary functions available in MRPATH: getImportanceSamples, sampleBetas, computeClusterMembProb. These functions are used after obtaining the MC-EM model fit to investigate SNP-specific latent variables in the MR-PATH model.
Plotting functions available in MRPATH: MRPATH_scatterplot, MRPATH_barplot.
You can find examples using the HDl-CHD dataset described in the preprint by running any of the following R commands.
library(MRPATH)
example("MR_PATH")
example("MRPATH_optimizeInitVals")
example("MRPATH_selectModel")
example("getImportanceSamples")
example("sampleBetas")
example("computeClusterMembProb")
example("MRPATH_scatterplot")
example("MRPATH_barplot")
Please report issues and suggestions for the MRPATH R package using the Github issues tracker.