Snakemake Batch Processing¶

VNtyper 2 includes a Snakemake workflow for processing multiple BAM files in parallel, with support for cluster execution.

Setup¶

The workflow file is located at snakemake/vntyper2.smk.

Create the Input File¶

List your BAM file paths in bams.txt, one per line:

/data/samples/sample1.bam
/data/samples/sample2.bam
/data/samples/sample3.bam

Each BAM filename (without .bam) becomes the sample name and output subdirectory name.

Create Output Directories¶

The workflow writes results to results/<sample_name>/ and logs to logs/<sample_name>.log.

Running Locally¶

snakemake -s snakemake/vntyper2.smk --use-conda -j 10

Cluster Execution¶

For SLURM or other cluster schedulers, use a Snakemake profile:

snakemake -s snakemake/vntyper2.smk --use-conda --profile slurm/

Resource Allocation¶

The workflow defines resource requests per job:

These are set in the run_vntyper_pipeline rule and can be adjusted by editing snakemake/vntyper2.smk.

Workflow Details¶

The workflow runs one rule per BAM file:

vntyper --config-path vntyper/config.json pipeline \
    --bam {input.bam} --threads 8 --reference-assembly hg38 \
    --fast-mode --keep-intermediates -o results/{sample}/

Output Structure¶

results/
├── sample1/
│   ├── kestrel/
│   │   └── kestrel_result.tsv
│   ├── pipeline_summary.json
│   └── ...
├── sample2/
│   └── ...
logs/
├── sample1.log
├── sample2.log

After batch processing, use the cohort analysis command to aggregate results:

vntyper cohort -i results/* -o cohort_output/