Running the analysis requires the Matlab-based scripting environment automaticanalysis (aa). A custom version of the code should be downloaded from the linked [GitHub repository][1]. Additionally, SPM12 (v7771), FSL (v6.0), and the rWLS and Wavelet despiking toolboxes must be installed.
A [Wiki][2] is available that describes aa setup.
Analysis scripts for each dataset are provided in the "Files" section of this page. These are standard Matlab mfiles. Customization of the data and results directory paths is necessary as indicated in the code. Also, the scripts assume the Matlab Parallel Computing Toolbox is available (and uses a "local" pool profile); if it is not, you should delete (or comment) the code block listed under "PCT customization" in the mfile.
The same processing is used for all but two of the datasets; ds001534 and ds002382 require a slightly modified pipeline (see the call to "aarecipe" in the analysis script). All three pipeline versions are provided.
Finally, some datasets required modifications to correct nifti or tsv irregularities -- see the Supplemental Materials for a description.
[1]: https://github.com/jones-michael-s/automaticanalysis
[2]: https://github.com/automaticanalysis/automaticanalysis/wiki