To use these scripts:

1. Download the Cambridge_Buckner data (parts 1 through 4) from
   fcon1000 under the NITRC site:

        http://www.nitrc.org/projects/fcon_1000

   or the newer:

        http://fcon_1000.projects.nitrc.org

2. Unpack the tar files under 'input_data'.  For example, there should
   end up being a directory tree for subject sub00156:

        input_data/Cambridge_Buckner/sub00156

   containing (at least):

        func/rest.nii.gz
        anat/mprage_skullstripped.nii.gz

3. Run the 4 top-level processing scripts, in order, one at a time:

        ./run.stage.1.preprocess.txt - through single subject regression
        ./run.stage.2.set.groups.txt - make motion groups
        ./run.stage.3.setup.GIC.txt  - prepare for group correlation test
        ./run.stage.4.run.GIC.txt    - run group correlation test

   The stage.1 preprocessing script can take a long time (analyzing 200
   subjects), and depends upon the number of CPUs that can be decicated.

   To alter the number of CPUs applied in the regression, change the 
   -jobs line of the afni_proc.py command in scripts/cmd.afni_proc .

- R Reynolds  11 Oct 2012