This package fits multi-outcome regression models to matched case-control and case-crossover data using Multi-Outcome Regression with Tree-structured Shrinkage (MOReTreeS).
MOReTreeS is a statistical method for analyzing the effect of an exposure on a large number of related outcomes. The model performs three important functions simultaneously: (1) discovers groups of outcomes that have a similar relationship with the exposure; (2) estimates group-specific exposure effects; (3) shares information between related outcomes about exposure effects and, where relevant, covariate effects. See the mortrees vignette for more information.
vignette("moretrees")
If using moretrees in published research, please cite the following paper:
Thomas EG, Trippa L, Parmigiani G, and Dominici F. Estimating the effects of fine particulate matter on 432 cardiovascular diseases using multi-outcome regression with tree-structured shrinkage. *Journal of the American Statistical Association*, 2020.
Currently, the package is only designed to fit MOReTreeS models to data that would normally be analyzed using conditional logistic regression, such as matched case-control or case-crossover data. However, MOReTreeS equivalents of other regression models are in development.
Moretrees can be installed by running the following code in your R session. Use the option build_vignettes = TRUE to access the moretrees vignette.
# install.packages("devtools")
library(devtools)
install_github("emgthomas/moretrees_pkg", build_vignettes = TRUE)
library(moretrees)
Emma Thomas.