Skip to content

Latest commit

 

History

History
99 lines (77 loc) · 3.8 KB

README.md

File metadata and controls

99 lines (77 loc) · 3.8 KB
Raw

Version DOI

Snakemake workflow: subcorticalparc-smk

Snakemake workflow for identifying structural and functional connectivity with regions/structures of interest.

This workflow is currently customized to run with data from the HCP1200 7T

Description

Using the HCP-MMP cortical parcellation (180 regions, left/right sym labels) as targets and performs probabilistic tracking from the region / structure of interest seed in each subject's native space. The connectivity data from seed voxels are brought into the template space to perform spectral clustering on the concatenated feature vectors to parcellate into k regions.

Inputs

  • Probabilistic segmentation(s) as 3D NIFTI for region/structure of interest on a single MNI template space
  • participants.tsv with target subject IDs
  • For each target subject:
    • Freesurfer processed data
    • Pre-processed DWI, registered to T1w space (e.g. HCP-style, or from prepdwi)
    • BEDPOST processed data
    • Transformations from template T1w space to/from each subject T1w, e.g. from: ants_build_template_smk; must include affine, warp and invwarp

Singularity containers

  • Freesurfer (for mri_convert, mris_convert, mri_info)
  • Connectome workbench
  • Ciftify
  • Neuroglia (contains FSL, ANTS, gnu parallel etc..)
  • FSL6 with CUDA container
  • pythondeps-zonaconn (dependencies for python rules)

NOTE: Currently the tractography step in the workflow requires a GPU

RECOMMENDED EXECUTION:

snakemake --profile cc-slurm

  • Further job grouping (beyond 1 job per subject) on graham with --group-components is not recommended as it can lead to over-subscribed GPUs when run on graham (TODO: fix this)

Authors

  • Ali Khan @akhanf
  • Sudesna Chakraborty
  • Jason Kai
  • Roy Haast

Reference

If you use this workflow in a paper, don't forget to give credits to the authors by citing the URL of this (original) repository and, if available, its DOI (see above).