This requires kslurm (https://github.com/pvandyken/kslurm) to be installed. You can do this by pasting this at a terminal:
curl -sSL https://raw.githubusercontent.com/pvandyken/kslurm/master/install_kslurm.py | python -
Tract map processing is best done using local scratch ($SLURM_TMPDIR) since it is very I/O intensive. Since there isn't an easy way to achieve this built-in to snakemake, another job script has been created that copies the data to local scratch, runs snakemake to generate a single subject's tractmaps, then copies that data back before the job finishes.
Before you run this job, however, you want to make sure:
- all the processing up to tract maps is complete. Can do this with:
#this command does a dry-run to see if any jobs need to be run still
snakemake --omit-from resample_clus_seed -npr
#this will run all the jobs to complete processing before tract maps
snakemake --omit-from resample_clus_seed --profile cc-slurm
- the
results/tractmapfolder for a subject should be empty submitting any new tractmap jobs.
To run tract maps for one subject (e.g. sub-100307) you can use:
kbatch 3:00 8 32G gpu -a ctb-akhanf_gpu . ./job_tractmaps sub-100307
To run on all subjects in a participants.tsv file, use:
./submit_tractmaps config/hcptestparticipants.tsv