/neuro-functional-pipeline

fMRIPrep containerized pipeline

Primary LanguageShell

Neuro functional pipeline

Repository documenting fMRIPrep usage within containerized environments such as Singularity

  1. Create a BIDS dataset for all subjects/sessions within a given project. Ideally, all subjects matching sequences with matching parameters. Each anat sequence should be defaced, and notes to this effect should exist in the readme.txt at project level. ./scripts/0_setup_bids.sh

  2. Run mriqc on each subject. Once all subjects are completed, run mriqc group (without SSIDS, it runs on the whole set automatically). ./scripts/1_run_mriqc_session.sh 001C ./scripts/2_run_mriqc_group.sh

  3. Run fmriprep on each subject (~10hrs with 5xCPU,4xCores,12GB mem) ./scripts/3_run_fmriprep_session.sh 001C

Singularity basic commands: shell exec run

Running jobs on Compute Canada

  1. currently re-writing code to be a single master script that will: a. upload a single-dataset raw-bids (given project, site, ssid) b. run/monitor/re-run fmriprep with doubled cpu/ram/time options if failed due to time-out c. download fmriprep results