XCP-D atlas error

Software Support
1

Summary of what happened:

Hi NeuroStars Community,
I’m running XCP-D on a single subject with fMRIprep output, but the command soon fails due to atlas errors. I would love your help to get to the bottom of this!

Command used (and if a helper script was used, a link to the helper script or the command generated):

singularity run --cleanenv \
  --bind /home/roey/cluster_shared/CFG/data/,\
/home/roey/cluster_shared/CFG/data/fmri/,\
/home/roey/cluster_shared/ \
  "/home/roey/cluster_shared/singularity_images/xcp_d-0.11.1.simg" \
  /home/roey/cluster_shared/CFG/data/fmri/derivatives \
  /home/roey/cluster_shared/CFG/data/fmri/derivatives/xcp_d_0_11_1_acompcor_withtedana_ortho_gsr \
  participant \
  -w /home/roey/cluster_shared/CFG/data/xcpd_work \
  --mode linc \
  --participant_label "sub-103" \
  --nuisance-regressors "/home/roey/cluster_shared/CFG/code/Automated_Pipeline/functional_pipeline/acompcor_tedana_gsr_xcpd_config.yaml" \
  --fs-license-file "/home/roey/cluster_shared/freesurfer/license.txt" \
  --datasets tedana="/home/roey/cluster_shared/CFG/data/fmri/derivatives/custom_confounds_for_xcpd_ortho" \
  --bids-filter-file "/home/roey/cluster_shared/CFG/code/Automated_Pipeline/functional_pipeline/xcpd_bids_filter.json"

Version:

0.11.1

Environment (Docker, Singularity / Apptainer, custom installation):

Singularity.

Data formatted according to a validatable standard? Please provide the output of the validator:

No validator.

Relevant log outputs (up to 20 lines):

Process Process-2:
Traceback (most recent call last):
  File "/usr/local/miniconda/lib/python3.10/multiprocessing/process.py", line 314, in _bootstrap
    self.run()
  File "/usr/local/miniconda/lib/python3.10/multiprocessing/process.py", line 108, in run
    self._target(*self._args, **self._kwargs)
  File "/usr/local/miniconda/lib/python3.10/site-packages/xcp_d/cli/workflow.py", line 94, in build_workflow
    retval['workflow'] = init_xcpd_wf()
  File "/usr/local/miniconda/lib/python3.10/site-packages/xcp_d/workflows/base.py", line 81, in init_xcpd_wf
    single_subject_wf = init_single_subject_wf(subject_id, anat_session, func_sessions)
  File "/usr/local/miniconda/lib/python3.10/site-packages/xcp_d/workflows/base.py", line 379, in init_single_subject_wf
    load_atlases_wf = init_load_atlases_wf()
  File "/usr/local/miniconda/lib/python3.10/site-packages/xcp_d/workflows/parcellation.py", line 58, in init_load_atlases_wf
    atlases = collect_atlases(
  File "/usr/local/miniconda/lib/python3.10/site-packages/xcp_d/utils/atlas.py", line 126, in collect_atlases
    if layout.get_dataset_description().get('DatasetType') != 'atlas':
  File "/usr/local/miniconda/lib/python3.10/site-packages/bids/layout/layout.py", line 997, in get_dataset_description
    return layouts[0].get_file('dataset_description.json').get_dict()
AttributeError: 'NoneType' object has no attribute 'get_dict'

Screenshots / relevant information:

This multi-band multi-echo data was processed with fMRIprep version 25.1.4, including tedana.

I also tried running XCP-D with an explicit setting of the atlases:

  --atlases "/usr/local/miniconda/lib/python3.10/site-packages/xcp_d/data/atlases"

but got the same error.

Finally, I opened the singularity image in a shell and verified that the atlases path exists and includes the appropriate dataset_desciption.json file.

2

Hi @roeysc what does the DatasetType field in that dataset_description.json say?

3

I think the missing dataset_description.json is probably in the tedana dataset. XCP-D needs to loop through all of the datasets that are fed in, including the automatically-included atlas datasets, to determine which ones contain atlases.

Does /home/roey/cluster_shared/CFG/data/fmri/derivatives/custom_confounds_for_xcpd_ortho have a dataset_description.json?

4

Thank you both, @Steven and @tsalo!
Great catch, @tsalo, indeed I was missing a dataset_description.json file in this directory. The atlas-related error threw me in the wrong direction. Thanks!