MRIQC fails with TraitError

winkler · April 12, 2019, 8:23pm

Hi Oscar,

Thanks again for updating. With v0.15.0 it is giving this other error…

Process Process-2:
Traceback (most recent call last):
  File "/usr/local/Anaconda/envs/py3.6/lib/python3.6/multiprocessing/process.py", line 258, in _bootstrap
    self.run()
  File "/usr/local/Anaconda/envs/py3.6/lib/python3.6/multiprocessing/process.py", line 93, in run
    self._target(*self._args, **self._kwargs)
  File "/gpfs/gsfs10/users/winkleram/ENIGMA/code/env/lib/python3.6/site-packages/mriqc/bin/mriqc_run.py", line 418, in init_mriqc
    wf_list.append(build_workflow(dataset[mod], mod, settings=settings))
  File "/gpfs/gsfs10/users/winkleram/ENIGMA/code/env/lib/python3.6/site-packages/mriqc/workflows/core.py", line 27, in build_workflow
    workflow = anat_qc_workflow(dataset, mod=mod, settings=settings)
  File "/gpfs/gsfs10/users/winkleram/ENIGMA/code/env/lib/python3.6/site-packages/mriqc/workflows/anatomical.py", line 87, in anat_qc_workflow
    norm = spatial_normalization(settings)
  File "/gpfs/gsfs10/users/winkleram/ENIGMA/code/env/lib/python3.6/site-packages/mriqc/workflows/anatomical.py", line 191, in spatial_normalization
    get_template(tpl_id, resolution=resolution, desc='brain', suffix='mask'))
  File "/gpfs/gsfs10/users/winkleram/ENIGMA/code/env/lib/python3.6/site-packages/nipype/interfaces/base/traits_extension.py", line 112, in validate
    self.info_text, value))
traits.trait_errors.TraitError: The trait 'reference_mask' of a RobustMNINormalizationInputSpecRPT instance is an existing file name, but the path  '[]' does not exist.

winkler · April 12, 2019, 9:00pm

Let me provide some extra piece of info: we don’t have .json sidercar files for this dataset, and this is only structural T1w.
At the top of the BIDS directory, in the same level as the dataset_description.json, we added a file named T1w.json that contains nothing but just an empty pair of curly braces, that is:

{ }

oesteban · April 13, 2019, 5:23am

What is the version of TemplateFlow?:

python -c 'import templateflow; print(templateflow.__version__)'

Next question is the output of:

python -c "from templateflow import api; print([str(f) for f in api.get('MNI152NLin2009c', resolution=2)])"

It is very likely that pip install -U templateflow would fix this issue, but those two commands will help to be sure.

winkler · April 13, 2019, 3:56pm

Templateflow is 0.1.7.

The 2nd output is an empty list:

$ python -c "from templateflow import api; print([str(f) for f in api.get('MNI152NLin2009c', resolution=2)])"
[]

oesteban · April 16, 2019, 4:48pm

Okay, I think I can guess what is happening. Your TemplateFlow 0.1.7 was not installed fresh, and that is confusing the machinery to work around git-annex. I’ve opened an issue in our repo. Thanks for reporting!

Let’s see if the following steps get you on your way while we find a better solution:

python -c "from templateflow import api; print(api.get('MNI152NLin2009cAsym', resolution=2, desc='brain', suffix='mask'))"

This should return [].

Then let’s try to fix the problem.

touch ${TEMPLATEFLOW_HOME:-$HOME/.cache/templateflow}/tpl-MNI152NLin2009cAsym/tpl-MNI152NLin2009cAsym_res-02_desc-brain_mask.nii.gz

Then, we re-run the first command again:

python -c "from templateflow import api; print(api.get('MNI152NLin2009cAsym', resolution=2, desc='brain', suffix='mask'))"

And now, after some download indicator, you should see that the output is '/some/path/tpl-MNI152NLin2009cAsym/tpl-MNI152NLin2009cAsym_res-02_desc-brain_mask.nii.gz'.

After these three steps, please post the output of the following command (I made a mistake in the original post), so that I can check whether you need to manually initiate more files for MRIQC to work:

python -c "from templateflow import api; print('\n'.join([str(f) for f in api.get('MNI152NLin2009cAsym', resolution=2)]))"

winkler · April 16, 2019, 5:48pm

Thanks Oscar!

The first returned [] indeed. Did then the touch and in the 3rd command it downloaded from amazonaws. The NIFTI file came, and it’s stored in ~/.cache/templateflow/tpl-MNI152NLin2009cAsym/tpl-MNI152NLin2009cAsym_res-02_desc-brain_mask.nii.gz as I believe it’s expected. The file has 29576 bytes.

But then the last command did not run successfully:

Traceback (most recent call last):
  File "<string>", line 1, in <module>
TypeError: 'PosixPath' object is not iterable

I made it iterable then transforming it to a list, hope this is fine:

python -c "from templateflow import api; print('\n'.join([str(f) for f in [api.get('MNI152NLin2009cAsym', resolution=2)]]))"

It gave then this output (except that instead of “~” it shows the full path):

~/.cache/templateflow/tpl-MNI152NLin2009cAsym/tpl-MNI152NLin2009cAsym_res-02_desc-brain_mask.nii.gz

Would it be ok if we tried to run MRIQC at this point? Thanks!

oesteban · April 16, 2019, 5:56pm

The fact that you only got one file from the (fixed) last command makes me worry.

Let’s start fresh to make sure MRIQC will work out.

rm -r ${TEMPLATEFLOW_HOME:-$HOME/.cache/templateflow}
python -c "from templateflow import api; print('\n'.join([str(f) for f in api.get('MNI152NLin2009cAsym')]))"

That should first clean up the template flow cache (beware of the rm command in the beginning, please proceed with caution -maybe checking with ls what is to be deleted-).

The second line should download the full MNI152NLin2009cAsym template and all related files and resolutions.

If the second command is successful and you get the following:

~/.cache/templateflow/tpl-MNI152NLin2009cAsym/CHANGES
~/.cache/templateflow/tpl-MNI152NLin2009cAsym/LICENSE
~/.cache/templateflow/tpl-MNI152NLin2009cAsym/template_description.json
~/.cache/templateflow/tpl-MNI152NLin2009cAsym/tpl-MNI152NLin2009cAsym_from-OASISTRT20_mode-image_xfm.mat
~/.cache/templateflow/tpl-MNI152NLin2009cAsym/tpl-MNI152NLin2009cAsym_res-01_desc-brain_mask.nii.gz
~/.cache/templateflow/tpl-MNI152NLin2009cAsym/tpl-MNI152NLin2009cAsym_res-01_desc-brain_T1w.nii.gz
~/.cache/templateflow/tpl-MNI152NLin2009cAsym/tpl-MNI152NLin2009cAsym_res-01_desc-BrainCerebellumExtraction_mask.nii.gz
~/.cache/templateflow/tpl-MNI152NLin2009cAsym/tpl-MNI152NLin2009cAsym_res-01_desc-BrainCerebellumRegistration_mask.nii.gz
~/.cache/templateflow/tpl-MNI152NLin2009cAsym/tpl-MNI152NLin2009cAsym_res-01_desc-carpet_dseg.nii.gz
~/.cache/templateflow/tpl-MNI152NLin2009cAsym/tpl-MNI152NLin2009cAsym_res-01_desc-eye_mask.nii.gz
~/.cache/templateflow/tpl-MNI152NLin2009cAsym/tpl-MNI152NLin2009cAsym_res-01_desc-face_mask.nii.gz
~/.cache/templateflow/tpl-MNI152NLin2009cAsym/tpl-MNI152NLin2009cAsym_res-01_desc-head_mask.nii.gz
~/.cache/templateflow/tpl-MNI152NLin2009cAsym/tpl-MNI152NLin2009cAsym_res-01_label-CSF_probseg.nii.gz
~/.cache/templateflow/tpl-MNI152NLin2009cAsym/tpl-MNI152NLin2009cAsym_res-01_label-GM_probseg.nii.gz
~/.cache/templateflow/tpl-MNI152NLin2009cAsym/tpl-MNI152NLin2009cAsym_res-01_label-WM_probseg.nii.gz
~/.cache/templateflow/tpl-MNI152NLin2009cAsym/tpl-MNI152NLin2009cAsym_res-01_PD.nii.gz
~/.cache/templateflow/tpl-MNI152NLin2009cAsym/tpl-MNI152NLin2009cAsym_res-01_T1w.nii.gz
~/.cache/templateflow/tpl-MNI152NLin2009cAsym/tpl-MNI152NLin2009cAsym_res-01_T2w.nii.gz
~/.cache/templateflow/tpl-MNI152NLin2009cAsym/tpl-MNI152NLin2009cAsym_res-02_desc-brain_mask.nii.gz
~/.cache/templateflow/tpl-MNI152NLin2009cAsym/tpl-MNI152NLin2009cAsym_res-02_desc-brain_T1w.nii.gz
~/.cache/templateflow/tpl-MNI152NLin2009cAsym/tpl-MNI152NLin2009cAsym_res-02_desc-carpet_dseg.nii.gz
~/.cache/templateflow/tpl-MNI152NLin2009cAsym/tpl-MNI152NLin2009cAsym_res-02_desc-DKT31_dseg.nii.gz
~/.cache/templateflow/tpl-MNI152NLin2009cAsym/tpl-MNI152NLin2009cAsym_res-02_desc-eye_mask.nii.gz
~/.cache/templateflow/tpl-MNI152NLin2009cAsym/tpl-MNI152NLin2009cAsym_res-02_desc-face_mask.nii.gz
~/.cache/templateflow/tpl-MNI152NLin2009cAsym/tpl-MNI152NLin2009cAsym_res-02_desc-fMRIPrep_boldref.nii.gz
~/.cache/templateflow/tpl-MNI152NLin2009cAsym/tpl-MNI152NLin2009cAsym_res-02_desc-head_mask.nii.gz
~/.cache/templateflow/tpl-MNI152NLin2009cAsym/tpl-MNI152NLin2009cAsym_res-02_label-CSF_probseg.nii.gz
~/.cache/templateflow/tpl-MNI152NLin2009cAsym/tpl-MNI152NLin2009cAsym_res-02_label-GM_probseg.nii.gz
~/.cache/templateflow/tpl-MNI152NLin2009cAsym/tpl-MNI152NLin2009cAsym_res-02_label-WM_probseg.nii.gz
~/.cache/templateflow/tpl-MNI152NLin2009cAsym/tpl-MNI152NLin2009cAsym_res-02_PD.nii.gz
~/.cache/templateflow/tpl-MNI152NLin2009cAsym/tpl-MNI152NLin2009cAsym_res-02_T1w.nii.gz
~/.cache/templateflow/tpl-MNI152NLin2009cAsym/tpl-MNI152NLin2009cAsym_res-02_T2w.nii.gz

MRIQC should be good to go.

winkler · April 16, 2019, 6:06pm

Great, got the same output, and have just submitted to run. Fingers crossed…!
Thanks!