Whole genome sequencing (WGS) analysis using nf-core/sarek pipeline and verifybamid

Whole-genome sequencing analysis with nf-core/sarek and VerifyBamID2 contamination QC on a Slurm cluster.

All scripts and commands should be run from the repository root because paths are relative.

Project structure

config/
  nextflow.example.config
  nextflow.local.config
  sarek.params.yaml
  verifybamid.params.yaml
  samplesheet.csv

scripts/
  run_sarek.sbatch
  submit_sarek.sh
  setup_verifybamid.sh
  verifybamid.nf
  run_verifybamid.sbatch
  submit_verifybamid.sh
  calculate_checksums.sh
  box_upload.sh
  download_gdrive_files.sh

results/
  sarek/
  verifybamid/

Requirements

• Nextflow, loaded through the cluster module system
• Slurm
• Singularity available inside submitted jobs
• curl for downloading VerifyBamID2 reference files

Configuration

Create a local Nextflow configuration:

cp config/nextflow.example.config config/nextflow.local.config

Edit config/nextflow.local.config for the local Slurm environment.

Optional .env settings:

SAREK_VERSION="3.9.0"
PROFILE="singularity"
NEXTFLOW_CONFIG="config/nextflow.local.config"
NXF_SINGULARITY_CACHEDIR="/path/to/singularity/cache"

Edit pipeline parameters in:

• config/sarek.params.yaml
• config/verifybamid.params.yaml

All paths in these files are relative to the repository root.

Run Sarek

./scripts/submit_sarek.sh

Outputs are written to:

results/sarek/

Run VerifyBamID2

VerifyBamID2 estimates human-to-human contamination from the final Sarek CRAM.

Edit the variables at the top of scripts/setup_verifybamid.sh, then prepare the input and reference files:

./scripts/setup_verifybamid.sh

Submit the workflow:

./scripts/submit_verifybamid.sh

Do not run the VerifyBamID2 Nextflow command directly on the login node when Singularity is only available inside submitted jobs.

Main outputs:

results/verifybamid/S001.selfSM
results/verifybamid/S001.Ancestry
results/verifybamid/S001.log

Inspect the contamination estimate:

column -t results/verifybamid/S001.selfSM

The relevant field is FREEMIX.

Helper scripts

• scripts/download_gdrive_files.sh: download FASTQ files from Google Drive
• scripts/calculate_checksums.sh: calculate checksums for files
• scripts/box_upload.sh: upload files to Box.com

Run each helper script from the repository root.

Monitor jobs

squeue -u "${USER}"
sacct -S today -u "${USER}" --format=JobID,JobName,State,ExitCode,Elapsed,Reason

Logs are stored under logs/.