SLURM Pipelines

Run PhenoTypic batch processing on SLURM-managed clusters.

Automatic SLURM Detection

When SLURM is available, the CLI automatically submits jobs to the scheduler. To force local execution instead:

python -m phenotypic --pipeline pipeline.json --input /plates/ -o /output/ --force-local

SLURM Arguments

Pass SLURM parameters with repeated --slurm flags:

python -m phenotypic --pipeline pipeline.json --input /plates/ -o /output/ \
    --slurm time=240 \
    --slurm slurm_partition=gpu \
    --slurm mem_gb=16

Wait for Completion

By default, the CLI returns immediately after submitting SLURM jobs. To wait:

python -m phenotypic --pipeline pipeline.json --input /plates/ -o /output/ --wait

Resume on SLURM

Resume works the same way as local execution:

python -m phenotypic --pipeline pipeline.json --input /plates/ -o /output/ --resume