This R tutorial is associated with the EPHOR deliverable report D4.1 on the “multiple exposure methods”. In that report, we gave a summary table of over 40 statistical approaches that may be useful for the exposome analyses and highlighted three methods that demonstrate potentially sufficient utility to EPHOR project for different analytical purposes: LASSO, Random forest, and BKMR (Bayesian Kernel Machine Regression).
This R tutorial aims to demonstrate how the three methods can be applied to sample data. We hope that this tutorial could support readers to apply the three methods to their analyses, however, because the three methods have been widely applied in many other fields, it is highly encouraged to also refer to other relevant sources of tutorials if necessary.
Here you will find:
R codes: EPHOR WP4_tutorial for lasso, rf, bkmr.Rmd
Dataset: data_2.xlsx
Codes with output EPHOR WP4_tutorial for lasso, rf, bkmr.html
Please feel free to get in touch if there are any suggestions for the tutorial via [email protected] or [email protected].