verywise

vertex-wise, whole-brain linear mixed models

The goal of verywise is to offer a flexible, user-friendly interface to whole-brain analysis of neuro-imaging data that has been pre-processed using FreeSurfer.

The package was specifically designed for the analysis of longitudinal (e.g. multi-session) and/or multi-site neuroimaging data.

Currently, verywise allows the estimation of vertex-wise Linear Mixed Models, but will be extended to other statistical models in the future.

It can handle imputed (phenotype) data from several packages (mice, mi, amelia, etc.).

Multiple testing correction is currently achieved using MCZ simulations from FreeSurfer. This means that you will need FreeSurfer installed and correctly set up.

Installation

You can install the development version of verywise from GitHub with:

# install.packages("devtools")
devtools::install_github("SereDef/verywise")

Basic use

library(verywise)

# Run a linear mixed model
run_vw_lmm(
  formula = vw_thickness ~ sex * age + site + (1 | id), # model formula
  subj_dir = "./VeryWiseUserUser/SimulatedExample", # Neuro-imaging data location
  hemi = "both", # (default) which hemispheres to run
  pheno = phenotype, # An R object already in memory or path to file 
)

You can find more info and tutorials here.

Visualization

To inspect and plot your results you can use our interactive web application, verywiseWIZard. You can run this locally or try it out here.

Note

This is a spin-off of the (more stable) QDECR package.

Funders

This work was supported by the FLAG-ERA grant Infant2Adult and by The Netherlands Organization for Health Research and Development (ZonMw, grant number 16080606).

All feedback and contributions are very welcome!.