WARNING!! A TR metadata error affected bandpass filtering in some RBC functional derivatives (NOT raw data or structural outputs). Impact varies by dataset TR (minimal for TR~1s, larger for TR>2s). Affected derivatives have been temporarily removed; corrected versions coming in phases over 3 months. For an immediate fix to C-PAC derivatives, you can apply this patch code. We have also added a new release for fMRIPrep + XCP-D derivatives as described below. Please see this link for a full description of this issue and our response.

07/06/2026: fMRIPrep and XCP-D Release


You can find the FreeSurfer, fMRIPrep and XCP-D derivatives in their corresponding repositories on GitHub.

The repositories (you can find them all here) are named as followed:

  • If you’re looking for fMRIPrep anatomical (processed with --anat-only) and FreeSurfer derivatives, the repo will be named <study>_fMRIPrep-Anat
  • If you’re looking for fMRIPrep functional derivatives, the repo will be named <study>_fMRIPrep-Func
  • If you’re looking for FreeSurfer-Post derivatives, the repo will be named <study>_FreeSurfer-Post
  • If you’re looking for XCP-D derivatives, the repo will be named <study>_XCP-D

where <study> is replaced with HBN, NKI, PNC, BHRC or CCNP.

Download data


Clone the repository, then use datalad get to retrieve file contents:

# Clone with HTTPS
datalad clone https://github.com/ReproBrainChart/<study>_<pipeline>.git

# Or clone with SSH
datalad clone git@github.com:ReproBrainChart/<study>_<pipeline>.git

# Example: clone the HBN XCP-D repository
datalad clone https://github.com/ReproBrainChart/HBN_XCP-D.git

# Then enter the repository
cd HBN_XCP-D

# Download specific files
datalad get sub-*/ses-*/func/*fsLR_den-91k_desc-denoised_bold.dtseries.nii

Boilerplate detailing processing steps is available in each repository at logs/CITATION.md.

fMRIPrep anat only command 
singularity run --cleanenv \
    -B "${PWD}" \
    -B "code/license.txt:/SGLR/FREESURFER_HOME/license.txt" \
    containers/.datalad/environments/fmriprep-<version>/image \
    inputs/data/BIDS \
    outputs/fmriprep_anat \
    participant \
    -w "${PWD}/.git/tmp/wkdir" \
    --stop-on-first-crash \
    --fs-license-file /SGLR/FREESURFER_HOME/license.txt \
    --output-spaces MNI152NLin6Asym:res-1 \
    --force-bbr \
    --skip-bids-validation \
    -vv \
    --anat-only \
    --cifti-output 91k \
    --n_cpus 9 \
    --mem-mb 14000 \
    --bids-filter-file "${session-filterfile}" \
    --participant-label "${subid}"
fMRIPrep func command 
singularity run --cleanenv \
    -B "${PWD}" \
    -B "code/license.txt:/SGLR/FREESURFER_HOME/license.txt" \
    containers/.datalad/environments/fmriprep-<version>/image \
    inputs/data/BIDS \
    outputs/fmriprep_func \
    participant \
    -w "${PWD}/.git/tmp/wkdir" \
    --stop-on-first-crash \
    --fs-license-file /SGLR/FREESURFER_HOME/license.txt \
    --output-spaces func T1w MNI152NLin6Asym:res-2 \
    --force-bbr \
    --skip-bids-validation \
    -vv \
    --cifti-output 91k \
    --n_cpus 4 \
    --mem-mb 30000 \
    --fs-subjects-dir inputs/data/fmriprep_anat/fmriprep_anat/sourcedata/freesurfer \
    --bids-filter-file "${session-filterfile}" \
    --participant-label "${subid}"
XCP-D command 
singularity run --cleanenv \
    -B ${PWD} \
    -B code/license.txt:/SGLR/FREESURFER_HOME/license.txt \
    containers/.datalad/environments/xcpd-0-10-6/image \
    inputs/data/fmriprep_func/fmriprep_func \
    outputs/xcpd \
    participant \
    -w ${PWD}/.git/tmp/wkdir \
    --stop-on-first-crash \
    --fs-license-file /SGLR/FREESURFER_HOME/license.txt \
    -vv \
    --mode linc \
    --dummy-scans auto \
    --input-type fmriprep \
    --nthreads 8 \
    --omp-nthreads 8 \
    --mem-gb 50 \
    --atlases 4S1056Parcels 4S156Parcels 4S256Parcels 4S356Parcels 4S456Parcels 4S556Parcels 4S656Parcels 4S756Parcels 4S856Parcels 4S956Parcels Glasser Gordon HCP MIDB Tian \
    --participant-label "${subid#sub-}"


The following QC files are available in each repository:

  • study-<study>_desc-T1_qc.tsv
  • study-<study>_desc-func_qc.tsv

Install DataLad

RBC is accessible via datalad. Follow the instructions here to get it installed.

Accessing RBC data via DataLad happens in two steps. You will first clone a RBC data repository from github. This will make a copy of the RBC file layout, but none of the actual data will be present. The next step is to get your data, which tells DataLad to download the content of specific files to your copy. Once the file content is present in your copy, you can use RBC data just like any other set of files.

Pick a dataset

You can find each of the unprocessed BIDS MRI data, processed functional and processed structural derivatives in their corresponding repositories on GitHub.

The repositories (you can find them all here) are consistently named such that:

  • If you’re looking for BIDS MRI, the repo will be named <study>_BIDS
  • If you’re looking for processed functional data, the repo will be named <study>_CPAC
  • If you’re looking for processed structural MRI, the repo will be named <study>_FreeSurfer

where <study> is replaced with HBN, NKI, PNC, BHRC or CCNP.

Note: Participant demographics and phenotypic data are under <study>_BIDS repository for each study and are named study-<study>_desc-participants.tsv.

Choose a QC threshold

When you download RBC data at a specific QC threshold you can be sure that any file you find in your downloaded data is usable for further analysis. RBC releases data at 3 different QC thresholds:

  • complete-pass: Structural and BOLD data have passed QC
  • complete-artifact: Structural data can be “Pass” or “Artifact”, BOLD is “Pass”
  • warning-fail: Contains all the RBC data including QC failures. This threshold is not recommended and you’ll have to explain why you chose this threshold in any publication resulting from it.

Choose a version

Lastly, you’ll need to pick a version of the data to get. Each time RBC makes changes to the data, it will be released with a new version number. By picking a version number for your RBC-based study you can be sure that your input data will never change. The current release of RBC is 0.1.

Clone the data

Getting data on to your system will involve running a command like this:

$ datalad clone \
    https://github.com/ReproBrainChart/<study>_<content>.git \
    -b <qc_threshold>-<version>

Example walkthrough

Suppose I’d like to get some processed structural data from PNC. I’ve decided that I don’t want to include T1ws with notable artifacts, so I choose the QC threshold to be complete-pass. I’ve checked the RBC website and see that the most recent release is 0.1. This means that my last argument to datalad clone will be -b complete-pass-0.1.

From the example command in ths documentation I replace <study> with PNC and <content> with FreeSurfer. My command would be:

$ datalad clone \
    https://github.com/ReproBrainChart/PNC_FreeSurfer.git \
    -b complete-pass-0.1

I see some warnings such as these:

[INFO   ] Unable to parse git config from origin
[INFO   ] Remote origin does not have git-annex installed; setting annex-ignore
[INFO   ] This could be a problem with the git-annex installation on the remote. Please make sure that git-annex-shell is available in PATH when you ssh into the remote. Once you have fixed the git-annex installation, run: git annex enableremote origin
install(ok): /home/cieslakm/data/PNC_FreeSurfer (dataset)

but these are ok. I now have a DataLad dataset with everything I need to start exploring RBC data.

That command finishes surprisingly quickly: PNC FreeSurfer has over a thousand subjects in it! But this is normal for DataLad. (Note that cloning the processed functional data (i.e., CPAC files) will take longer as CPAC functional outputs include more files compared to FreeSurfer.) There is now a PNC_FreeSurfer directory that I can look around in. Since I don’t know which files I might want, I will explore this directory in my terminal:

$ cd PNC_FreeSurfer/freesurfer/sub-192413932
$ ls
sub-192413932_brainmeasures.json@  sub-192413932_fsaverage.tar.xz@
sub-192413932_brainmeasures.tsv@   sub-192413932_fsLR_den-164k.tar.xz@
sub-192413932_freesurfer.tar.xz@   sub-192413932_regionsurfacestats.tsv@

These two tsv files contain the anatomical variables in tabular form, so I’d like to copy these contents to my computer:

$ datalad get *.tsv
get(ok): freesurfer/sub-192413932/sub-192413932_regionsurfacestats.tsv (file) [from output-storage...]
get(ok): freesurfer/sub-192413932/sub-192413932_brainmeasures.tsv (file) [from output-storage...]
action summary:
  get (ok: 2)

Shell glob patterns can be used to get whichever files you might need.

FAQs

Running datalad get appears to be stuck - what should I do?

The first time you run datalad get may take 30 minutes or more to finish copying the file. Subsequent runs of datalad get will be much faster.

Unless you see an error message when running datalad get, you should let the command run until you see something printed out.

Do I need a supercomputer to get this data?

You can clone any of these datasets to any machine with internet access, DataLad (git and git-annex) installed and your agreement with the License. The rest of the answer depends on what type of data you plan on copying to your computer:

  • If you plan on copying only tabular data content to your computer you will need at most a few free GB.
  • If you plan on copying 3D imaging data (e.g. ALFF, REHO) you will need hundreds of GB free.
  • If you plan on copying 4D denoised BOLD timeseries data you will need up to 4TB of free space.