nf-core pipeline · annotation, cancer, gatk4
nf-core/sarek
Nextflow pipeline for detecting germline and somatic variants from whole-genome and targeted sequencing data.
What nf-core/sarek is
nf-core/sarek is a comprehensive bioinformatics workflow designed to detect germline and somatic genetic variants from whole-genome sequencing and targeted sequencing assays, including whole-exome sequencing and targeted gene panels. Initially developed by Maxime U Garcia and Szilveszter Juhos at the National Genomics Infrastructure and National Bioinformatics Infrastructure Sweden at SciLifeLab, with backing from The Swedish Childhood Tumour Biobank, it has expanded through contributions from QBiC and the broader nf-core community.
Analysing high-throughput DNA sequencing requires coordinating multiple computational steps, including quality assessment, read alignment, deduplication, base quality score recalibration, variant discovery across several mutational classes, and downstream functional annotation. Sarek unifies these tasks into an automated pipeline capable of processing single-sample germline workflows as well as matched tumour-normal pairs, including complex longitudinal profiles with multiple relapses. Although initially built for human and mouse analyses, the pipeline can be applied to any organism with an assembled reference genome.
What the pipeline does
Sarek ingests sequencing data through a comma-separated samplesheet mapping patient, sample, and lane metadata to paired-end FASTQ files. The preprocessing phase supports optional consensus read generation from unique molecular identifiers (UMIs) using fgbio, adapter and quality trimming via FastQC and fastp, and contaminant sequence filtering using BBSplit. Sequence reads are aligned against a reference genome using BWA-mem, BWA-mem2, dragmap, or Sentieon BWA-mem, followed by coordinate sorting, duplicate marking, and base quality score recalibration using either GATK tools (MarkDuplicates, BaseRecalibrator, ApplyBQSR) or Sentieon alternatives (LocusCollector, Dedup). Alignment depth and coverage statistics are generated using samtools stats and mosdepth.
Variant calling is executed using a modular selection of tools covering single-nucleotide variants, short insertions and deletions, copy number alterations, structural variants, and microsatellite instability. Supported callers include GATK HaplotypeCaller, GATK Mutect2, DeepVariant, Strelka, Manta, FreeBayes, LoFreq, MuSE, TIDDIT, Control-FREEC, CNVkit, ASCAT, and MSIsensor variants. Following calling, variant call format (VCF) files can be filtered, normalised, and merged into consensus sets using bcftools or evaluated with Varlociraptor. Finally, functional annotation can be added using Ensembl VEP, SnpEff, SnpSift, or BCFtools annotate, and aggregate run statistics are collated in a MultiQC report.
Key facts
- Workflow framework
- Nextflow DSL2
- Container engines
- Docker, Singularity
- Input formats
- CSV samplesheet, paired-end FASTQ
- Sequencing types
- Whole-genome sequencing (WGS), targeted sequencing, whole-exome sequencing (WES)
- Variant classes
- Germline and somatic SNVs, indels, CNVs, SVs, and microsatellite instability
- Lead maintainers
- Friederike Hanssen and Maxime U Garcia
- Latest release
- 3.10.0
What you can use it for
- Somatic cancer variant profiling
- Identifies somatic single-nucleotide variants, indels, structural variants, and copy number alterations across matched tumour-normal pairs and multi-relapse patient series.
- Germline variant detection
- Discovers inherited single-nucleotide variants and small indels from whole-genome or targeted exome sequencing data using callers such as DeepVariant, GATK HaplotypeCaller, or Sentieon Haplotyper.
- Multi-caller consensus variant calling
- Runs multiple independent variant calling algorithms in parallel and derives consensus call sets through bcftools intersection or Varlociraptor filtering.
Run it here
The research agent runs nf-core/sarek with Nextflow on its own machine — up to 32 cores, 256 GiB of memory and 24 hours per run, which covers the test profile and real single-sample runs. It builds the samplesheet, checks it against the pipeline's schema and reads the results back. When a run is too big for one machine, it gives you the exact command for your own cluster instead.
Strengths and limits
Good at
- Supports a wide array of variant callers addressing single-nucleotide variants, small indels, structural variants, copy number changes, and microsatellite instability.
- Functions across any organism with an available reference genome, beyond its original human and mouse targets.
- Standardises end-to-end processing from raw reads through post-call filtering and functional annotation within containerised environments.
- Benchmarked continuously on releases against full-sized datasets, including Genome-in-a-Bottle NA12878 and SEQ2C consortium samples.
Watch out for
- GPU-accelerated Parabricks alignment and HaplotypeCaller execution remain experimental features.
- Pipeline parameters cannot be supplied through custom Nextflow configuration files and must be specified via the command-line interface or a parameter file.
Questions
Can I run nf-core/sarek without my own cluster?
The research agent runs nf-core/sarek with Nextflow on its own machine — up to 32 cores, 256 GiB of memory and 24 hours per run, which covers the test profile and real single-sample runs. It builds the samplesheet, checks it against the pipeline's schema and reads the results back. When a run is too big for one machine, it gives you the exact command for your own cluster instead.
What licence is nf-core/sarek under?
MIT, like every nf-core pipeline.
Can Sarek analyse species other than humans?
Yes. While initially designed for human and mouse genetics, Sarek can process sequencing data from any species that has a reference genome available.
Does Sarek support tumour data from multiple disease stages?
Yes. Sarek can handle paired tumour and normal samples as well as series that include additional relapse samples from the same patient.
How should input files be formatted for the pipeline?
Inputs must be listed in a CSV samplesheet specifying patient, sample, and sequencing lane identifiers alongside paths to paired-end FASTQ files.
Which variant callers are integrated into Sarek?
Sarek includes ASCAT, CNVkit, Control-FREEC, DeepVariant, FreeBayes, GATK HaplotypeCaller, GATK Mutect2, indexcov, LoFreq, Manta, mpileup, MSIsensor2, MSIsensor-pro, MuSE, Sentieon Haplotyper, Strelka, and TIDDIT, as well as an experimental Parabricks HaplotypeCaller.
How to cite it
Friederike Hanssen, Maxime U Garcia, Lasse Folkersen, Anders Sune Pedersen, Francesco Lescai, Susanne Jodoin, Edmund Miller, Oskar Wacker, Nicholas Smith, nf-core community, Gisela Gabernet, Sven Nahnsen Scalable and efficient DNA sequencing analysis on different compute infrastructures aiding variant discovery NAR Genomics and Bioinformatics Volume 6, Issue 2, June 2024, lqae031, doi: 10.1093/nargab/lqae031.
Written from nf-core/sarek's own README. Statements citing a figure that does not appear there are removed before publication. Licence and running details come from our registry.