[ENH] add QSIPrep

[michellewang]

• Closes #248

Work in progress!

• [x] Descriptor file
• [x] Default invocation file
• [ ] Default tracker config file
• [x] Update sample global config files

[codecov[bot]]

Codecov Report

All modified and coverable lines are covered by tests :white_check_mark:

Project coverage is 98.64%. Comparing base (fc3971f) to head (f1c5704).

Additional details and impacted files

```diff @@ Coverage Diff @@ ## main #266 +/- ## ======================================= Coverage 98.64% 98.64% ======================================= Files 28 28 Lines 1766 1766 ======================================= Hits 1742 1742 Misses 24 24 ```

:umbrella: View full report in Codecov by Sentry.
:loudspeaker: Have feedback on the report? Share it here.

[michellewang]

Added a basic invocation file which creates the following command (most of these are QSIPrep defaults)

Generated Command:
[[NIPOPPY_CONTAINER_COMMAND]] [[NIPOPPY_FPATH_CONTAINER]] qsiprep \
    '[[NIPOPPY_DPATH_BIDS]]' '[[NIPOPPY_DPATH_PIPELINE_OUTPUT]]' participant \
    --skip_bids_validation --participant_label '[[NIPOPPY_PARTICIPANT_ID]]' \
    --bids-database-dir '[[NIPOPPY_DPATH_PIPELINE_BIDS_DB]]' --recon-input-pipeline qsiprep \
    --omp-nthreads 4 -v -v --anat-modality T1w --b0-threshold 100 --dwi_denoise_window auto \
    --denoise-method dwidenoise --b1-biascorrect-stage final --distortion-group-merge none \
    --anatomical-template MNI152NLin2009cAsym --output-resolution 1.2 --b0-to-t1w-transform Rigid \
    --intramodal-template-transform BSplineSyN --b0-motion-corr-to iterative --hmc-transform Affine \
    --hmc_model eddy --shoreline_iters 2 --skull-strip-template OASIS \
    --fs-license-file '[[FREESURFER_LICENSE_FILE]]' --pepolar-method TOPUP \
    --fmap-no-demean -w '[[NIPOPPY_DPATH_PIPELINE_WORK]]'

Notes

• QSIPrep does not have a "session" argument, and with multi-session studies where sessions are processed separately (like we do) there might be overwriting of files in sub-<PARTICIPANT_ID>/anat, see https://github.com/PennLINC/qsiprep/issues/757. Not sure if that is a problem for us, but if it is then we might have to do an outer ses-<SESSION_ID> directory like we do for Freesurfer
• There is an argument freesurfer_input: we probably want to point this to the Freesurfer output from fMRIPrep to avoid running Freesurfer twice. To be tested I think