Mackey Lab CUBIC Data Organization - mackeylab/home GitHub Wiki

This information is current as of June, 2023.

Mackey Lab project data directory: /cbica/projects/cbpd_main_data

As a Mackey Lab member, you will have your own username & account that is part of the cbpdmaindata group on CUBIC. Your home directory will be located at /cbica/home/username. The lab’s project directory is located at /cbica/projects/cbpd_main_data.

When working in the project directory, you can either use your personal account, or use sudo to “become” the cbpdmaindata user. Information on the use of sudo on CUBIC is available here. To become the project user, once already logged into CUBIC, run the command: sudo -u cbpdmaindata sudosh or sudo -u cbpdmaindata bash You will then be prompted to enter your own password.

The benefit of working as the project user is that it keeps file and directory ownership consistent, so that we do not run into permission problems. For example, all BIDS preprocessing is done as the cbpdmaindata project user so that everyone who is part of our group can access these files. It is not possible to use ssh to mount the cluster locally as the cbpdmaindata user.

Important subdirectories within cbpd_main_data:

dicoms/

  • Raw dicom files copied from Flywheel to the cluster. These files are then used in the BIDS preprocessing pipeline.

CBPD_bids/

  • [see README file in this directory]
  • Contains all Mackey Lab neuroimaging data processed in BIDS format.
  • At the top level, each subject’s directory is named with the conventional format sub-CBPDXXXX, and longitudinal data is nested within each subject under ses-01, ses-02, ses-03, etc.
  • Within this directory, the derivatives/ directory contains the output of processing tools such as fmriprep and freesurfer. The derivative data is generally organized by longitudinal timepoint. For example, freesurfer_t1/ contains the freesurfer output for all baseline scans, while freesurfer_t2 contains the freesurfer output for time 2 scans.
  • The directories freesurfer_t1, freesurfer_t2, and freesurfer_t3 contain the raw, unedited freesurfer output for all subjects. The directories freesurfer_edits_t1, freesurfer_edits_t2, and freesurfer_edits_t3 contain the freesurfer output that has been manually edited to correct the surfaces as needed. Subjects are copied into the freesurfer_edits_ directories as surface editing is completed.

CBPD_bids_crosssectional/

  • [see README file in this directory]
  • Contains Mackey Lab neuroimaging data processed cross-sectionally in 2019/2020. Longitudinal subjects are named with their subject ID plus the number of the longitudinal visit at the end (e.g., sub-CBPD0126 is the first scan visit for this subject, sub-CBPD01262 is the second scan visit for the same subject). Newer subjects that have been scanned since 2020 are not included in this directory. BIDS data is now solely written to the CBPD_bids/ directory.

NWBD_brannonlab/ and TEMP_NWBD/

  • NWBD is the name of the study from Liz Brannon’s lab that collected some of the same sequences in the same age range as our CBPD study, so we sometimes include NWBD subjects in our analyses. Data from this study is located in these two directories.
  • Note that processed NWBD data also exists in some other directories. For example, the freesurfer output exists in the CBPD_bids/derivatives/freesurfer_ directories.

piper/

  • Contains processed data and analyses of the Piper scans (functional scans while kids are watching Pixar’s “Piper” short film).

surfaces_editing/

  • Contains freesurfer output for subjects that need to be manually edited, or are in the process of manual editing. Once surface editing is complete, subjects are copied into the corresponding freesurfer_edits directory (e.g. CBPD_bids/derivatives/freesurfer_edits_t1).
  • This directory is only used for manual surface editing and should not be used for structural analyses.

code/

  • Contains scripts for various processing steps and analysis pipelines. Importantly, contains the scripts used to run BIDS preprocessing in the directory bids_ppc_scripts/.

qsub_output/

  • Directory where qsub output files are often written to. For example, all of the BIDS preprocessing scripts specify writing the output of the job log to this directory. (If no output location is specified, qsub will automatically write the log file to the same location as the script that it is running).