What are the expected outputs of dual regression preprocessing?

1

I am trying to find which output to use in a group analysis of the dual regression Smith et al 10 large scale networks.

pipeline config:

seed_based_correlation_analysis:
  run: true
  sca_roi_paths:
    /cpac_templates/PNAS_Smith09_rsn10.nii.gz: DualReg

No errors were logged, but I do not seem spatial map images in the outputs:

./105105-100_3_ses-1/func/sub-105105-100_3_ses-1_task-func-1_desc-mean-1_bold.nii.gz
./105105-100_3_ses-1/func/sub-105105-100_3_ses-1_task-func-1_desc-mean-2_bold.nii.gz
./105105-100_3_ses-1/func/sub-105105-100_3_ses-1_task-func-1_desc-preproc-1_bold.nii.gz
./105105-100_3_ses-1/func/sub-105105-100_3_ses-1_task-func-1_desc-preproc-2_bold.nii.gz
./105105-100_3_ses-1/func/sub-105105-100_3_ses-1_task-func-1_from-bold_to-template_mode-image_desc-1_xfm.nii.gz
./105105-100_3_ses-1/func/sub-105105-100_3_ses-1_task-func-1_from-bold_to-template_mode-image_desc-2_xfm.nii.gz
./105105-100_3_ses-1/func/sub-105105-100_3_ses-1_task-func-1_from-template_to-bold_mode-image_desc-1_xfm.nii.gz
./105105-100_3_ses-1/func/sub-105105-100_3_ses-1_task-func-1_from-template_to-bold_mode-image_desc-2_xfm.nii.gz
./105105-100_3_ses-1/func/sub-105105-100_3_ses-1_task-func-1_space-bold_desc-brain-1_mask.nii.gz
./105105-100_3_ses-1/func/sub-105105-100_3_ses-1_task-func-1_space-bold_desc-brain-2_mask.nii.gz
./105105-100_3_ses-1/func/sub-105105-100_3_ses-1_task-func-1_space-T1w_desc-mean-1_bold.nii.gz
./105105-100_3_ses-1/func/sub-105105-100_3_ses-1_task-func-1_space-T1w_desc-mean-2_bold.nii.gz
./105105-100_3_ses-1/func/sub-105105-100_3_ses-1_task-func-1_space-template_desc-bold-1_mask.nii.gz
./105105-100_3_ses-1/func/sub-105105-100_3_ses-1_task-func-1_space-template_desc-bold-2_mask.nii.gz
./105105-100_3_ses-1/func/sub-105105-100_3_ses-1_task-func-1_space-template_desc-mean-1_bold.nii.gz
./105105-100_3_ses-1/func/sub-105105-100_3_ses-1_task-func-1_space-template_desc-mean-2_bold.nii.gz
./105105-100_3_ses-1/func/sub-105105-100_3_ses-1_task-func-1_space-template_desc-preproc-1_bold.nii.gz
./105105-100_3_ses-1/func/sub-105105-100_3_ses-1_task-func-1_space-template_desc-preproc-2_bold.nii.gz
./105105-100_3_ses-1/func/sub-105105-100_3_ses-1_task-func-1_space-template_res-derivative_desc-bold-1_mask.nii.gz
./105105-100_3_ses-1/func/sub-105105-100_3_ses-1_task-func-1_space-template_res-derivative_desc-bold-2_mask.nii.gz

Am I misunderstanding something?

2

Hi Joshua,

There’s actually a recently-discovered bug in pipeline-config parsing that is reverting any changes to seed_based_correlation_analysis to the default (which includes run: Off), so to answer your questions:

What are the expected outputs of dual regression preprocessing?

Files with desc-DualReg in the name ending with _correlations and _statmap

Am I misunderstanding something?

Nope, just hitting a bug! It will be fixed in the shortly forthcoming C-PAC v1.8.5.

If you can’t wait a few weeks for the fix and are comfortable using a less-than-fully-tested development image, docker://ghcr.io/fcp-indi/c-pac:fix_minimal-schema includes the proposed fix for this bug.

I’ll post back here when 1.8.5 is released with the fix included.

3

Apologies for leaving this thread hanging for months past the release of C-PAC 1.8.5 ― there’s an unresolved error in testing the pip-installable cpac package, so the 1.8.5-compatible version (v0.6.0) hasn’t been released, though you can install the development version of the wrapper (if you use the wrapper) directly from GitHub like

pip install git+https://github.com/FCP-INDI/cpac.git@main

If you use C-PAC through Docker or Apptainer / Singularity directly, 1.8.5 should already work for you as expected, and this issue should be resolved.