Neuropointillist Tutorial

Neuropointillist logo

This is a tutorial for using neuropointillist to run some examples. This assumes that you have already followed the directions in Installation.

Setting your PATH variable

After you have downloaded and installed the neuropointillist programs in a directory, you need to add this directory to your PATH variable. Suppose that you have downloaded the neuropointillist package into ~/neuropointillist. Assuming you are running the bash shell, edit your PATH as follows:

export PATH=$PATH:~/neuropointillist

This code will make it so that when you type npoint or npointrun at the command line, your shell can find them. You can put this in your ~/.bashrc file (if you use bash) so that you don’t have to do this every time you log in.

For detailed usage information, see the Usage. If you just want to get going, read on!

Simple fMRI example

This is a quick start tutorial that uses simulated fMRI data in the directory example.rawfmri. This explains how to set up a model, debug it, and run it on raw fMRI data. This is primarily for educational purposes, because fMRI software is a lot faster and better at doing this).

Advanced fMRI example

This tutorial uses a reproducibility data set downloaded from OpenNeuro to illustrate some of the more interesting things one might do in R that are difficult to achieve with fMRI software.

Permutation testing fMRI example

This tutorial uses the data set and results from the advanced example, above, to demonstrate how you can use Neuropointillist to implement permutation testing approaches to cluster-based corrections for multiple comparisons (specifically, Equitable Thresholding and Clustering in AFNI).

Flournoy example (longitudinal)

This tutorial illustrates the kinds of analyses that form the primary motivation for neuropointillist: the ability to compare forms of change and run SEM growh models on single-subject data that has completed first level processing.

Correcting for multiple comparisons

This tutorial provides some guidance for how to correct for multiple comparisons, primarily in the case where you are working with processed first level data.