Snakemake Workflow¶

For batch processing multiple VCF files in parallel, hardnormly includes a Snakemake 8+ workflow with SLURM support.

Overview¶

The workflow reads a list of VCF file paths and runs hardnormly.sh on each one independently. Configuration (reference paths, BED files, filters) is shared across all samples via config.yaml.

Setup¶

Edit configuration: Adjust config/config.yaml with your reference paths, BED files, and filter settings.
Create VCF list: Create a text file with one VCF path per line. Set paths.vcf_list in the config to point to this file.
Choose profile: Use profiles/default/ for local execution or create a cluster profile (see profiles/charite/ for a SLURM example).

Running¶

# Dry run — preview what will be executed
snakemake --snakefile workflow/Snakefile --configfile config/config.yaml \
  --workflow-profile profiles/default -n

# Local execution
snakemake --snakefile workflow/Snakefile --configfile config/config.yaml \
  --workflow-profile profiles/default

# SLURM submission (auto-detects cluster environment)
sbatch scripts/run_snakemake.sh

Workflow Structure¶

workflow/
├── Snakefile                # Entry point (config validation, rule imports)
├── rules/
│   ├── common.smk           # Config shortcuts and helper functions
│   └── hardnormly.smk       # Pipeline rule (calls hardnormly.sh per sample)
├── envs/
│   └── hardnormly.yaml      # Lightweight conda env for the pipeline rule
└── schemas/
    └── config.schema.yaml   # JSON Schema for config validation

config/
└── config.yaml              # Workflow configuration

profiles/
├── default/
│   └── config.yaml          # Default resources and execution settings
└── charite/
    └── config.yaml          # Charite SLURM cluster settings

Configuration¶

Key sections in config/config.yaml:

Section	What it controls
`ref`	Reference genome FASTA and genome build
`paths`	Input VCF list, output directory, log directory
`regions`	Include/exclude BED files, slop value
`filtering`	Filter file path, PASS-only flag
`processing`	Stats generation, auto-indexing, plot generation

Config is validated against workflow/schemas/config.schema.yaml at workflow start. Invalid configuration fails fast with a clear error message.

Profiles¶

Profiles control resource allocation (threads, memory, runtime) without modifying rules:

profiles/default/ — Sensible defaults for local execution
profiles/charite/ — SLURM cluster settings (partition, account, walltime)

Create your own profile by copying profiles/default/ and adjusting values for your cluster.

Environment¶

The workflow uses Snakemake's --use-conda feature. Only a snakemake (8+) base environment is needed — per-rule conda environments are created automatically from workflow/envs/hardnormly.yaml.