This repository contains all R Markdown files used for the data analysis presented in Cardner et al. (2020), doi: 10.1172/jci.insight.131491.
For reasons of data privacy, this repository contains no actual data, but merely the code used to analyse it. The raw data was collated and stored as tidy data in RDS files under computations/ (not committed). In general, the data frames were stored in a long format, with a column named key coding for the feature (clinical, protein, lipid) and a column named value recording the measurement. A column named Proband ID contained an anonymised subject code.
- Normalisation
- Proteomics and lipidomics in
OmicsNormalisation.Rmd - NMR spectroscopy in
Nightingale.Rmd - Functional bioassays in
FunctionalNormalisation.Rmd - Clinical imputation in
ClinicalImputation.Rmd
- Proteomics and lipidomics in
- Mixed-effects modelling in
MixedEffects.Rmd - Logistic regression in
glmnet.Rmd - PLS regression in
PLS.Rmd - Graphical lasso in
HugeNPN.Rmd - Helper functions
helper_functions.Rhelper_normalisation.Rhelper_volcanoes.R
In addition to the main documents, there are auxiliary R Markdown files for various purposes. Importantly, the JointGGM.Rmd explores the joint graphical lasso.