nf-core pipeline ·
nf-core/rnadnavar
Somatic mutation detection pipeline integrating RNA and DNA sequencing data using a consensus calling approach.
What nf-core/rnadnavar is
nf-core/rnadnavar is a bioinformatics analysis pipeline designed for detecting somatic mutations by integrating RNA and DNA sequencing data. Built for cancer research, it was originally developed by Raquel Manzano Garcia at the Cancer Research UK Cambridge Institute with support from Maxime U Garcia. The workflow is based on RNA-MuTect, which was originally published by Yizhak et al. in 2019.
Calling somatic variants from RNA sequencing presents distinct challenges, particularly elevated background noise compared to genomic DNA. nf-core/rnadnavar addresses this by combining multiple variant calling algorithms into a consensus approach and applying targeted realignment and filtering to deliver annotated somatic variants.
What the pipeline does
The pipeline processes raw paired-end sequencing reads or pre-aligned data. When starting from raw reads, it performs quality control and adapter trimming using FastQC and fastp, followed by reference mapping using BWA-mem, BWA-mem2, dragmap, or STAR. Preprocessing for both DNA and RNA bulk sequencing relies on GATK modules, including MarkDuplicates, SplitNCigarReads, BaseRecalibrator, and ApplyBQSR, with coverage and alignment statistics collected via samtools stats and mosdepth.
Variant calling is executed using Mutect2, Strelka2, and SAGE. Variant call format (VCF) files are normalised with VT and converted to Mutation Annotation Format (MAF) to perform consensus calling across tools. An RNA-specific realignment stage extracts and re-processes reads in variant regions to account for RNA background noise. Variants are subsequently annotated using VEP and filtered using gnomAD frequencies, blacklists, whitelists, and RNA-specific filtering criteria.
Key facts
- Latest release
- 1.0.0 (2026-06-29)
- Original author
- Raquel Manzano Garcia (Cancer Research UK Cambridge Institute)
- Workflow framework
- Nextflow DSL2
- Supported aligners
- BWA-mem, BWA-mem2, dragmap, STAR
- Supported variant callers
- Mutect2, Strelka2, SAGE
- Accepted input data
- Paired-end FASTQ, BAM, CRAM, or MAF/VCF files
- Annotation tool
- VEP (Variant Effect Predictor)
What you can use it for
- Integrated somatic variant calling
- Detect somatic mutations from matched tumour RNA and normal DNA sequencing samples.
- Consensus somatic analysis across multiple callers
- Combine predictions from Mutect2, Strelka2, and SAGE to generate a consensus set of filtered MAF annotations.
- RNA variant refinement
- Extract reads overlapping candidate variant regions and realign them to remove false positives caused by RNA background noise.
Run it here
The research agent runs nf-core/rnadnavar 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
- Combines predictions across Mutect2, Strelka2, and SAGE into a unified consensus mutation set.
- Includes dedicated realignment and specialised filtering to mitigate false positives driven by RNA background noise.
- Supports flexible starting points, from raw FASTQ reads to pre-aligned BAM or CRAM files and existing MAF files.
- Uses DSL2 containerisation with one Docker or Singularity container per process.
Watch out for
- Every RNA tumour sample must be paired with a corresponding normal DNA sample from the same patient identifier.
- Only paired-end reads are supported when starting from raw FASTQ files.
Questions
Can I run nf-core/rnadnavar without my own cluster?
The research agent runs nf-core/rnadnavar 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/rnadnavar under?
MIT, like every nf-core pipeline.
Is a tumour DNA sample mandatory to run the pipeline?
No. A tumour DNA sample is optional, but each tumour RNA sample must be paired with a normal DNA sample from the same patient.
Can pre-aligned BAM or CRAM files be provided as inputs?
Yes. The pipeline can accept pre-aligned BAM or CRAM files along with their respective index files instead of raw FASTQ reads.
How are sample types designated in the input samplesheet?
Sample types are defined in the status column using 0 for normal DNA, 1 for tumour DNA, and 2 for tumour RNA.
Can the pipeline start directly from MAF files?
Yes. If skipping realignment, the workflow can start at the consensus stage using MAF files and their caller annotations.
How to cite it
10.5281/zenodo.21038367
Written from nf-core/rnadnavar's own README. Statements citing a figure that does not appear there are removed before publication. Licence and running details come from our registry.