nf-core pipeline · atac-seq, chip-seq, deseq2
nf-core/differentialabundance
Performs differential abundance and gene set analyses on observation matrices from RNA-seq and microarrays.
What nf-core/differentialabundance is
nf-core/differentialabundance is a bioinformatics pipeline designed to analyse observation matrices across experimental groups, producing differential statistics, diagnostic visualisations, and downstream functional evaluations. Originally developed by Jonathan Manning and Oskar Wacker, it caters to high-throughput comparative datasets such as RNA sequencing count tables and Affymetrix microarray data.
The pipeline standardises post-quantification workflows by validating input consistency between feature matrices, experimental contrasts, and sample annotations. By automating differential testing, exploratory plotting, gene set enrichment, and interactive reporting, it allows researchers to systematically inspect comparative statistics across multiple experimental conditions.
What the pipeline does
The pipeline begins by validating the relationship between the input observation matrix, sample annotations, and specified contrast definitions, optionally extracting genomic feature annotations from a reference GTF file if one is provided. It then executes differential abundance statistical testing across each defined contrast using frameworks such as DESeq2 or limma-voom. When gene set annotations are supplied in GMT format, differential gene set analysis is conducted alongside the primary contrast calculations.
Following differential testing, the workflow outputs exploratory and differential visualisations and compiles an HTML report driven by Quarto markdown and shinyngs. It packages a parameterised Quarto notebook zip archive for offline modification in RStudio and can optionally build an interactive Shiny data-mining application that runs locally or deploys to shinyapps.io.
Key facts
- Pipeline framework
- Nextflow DSL2
- Latest release
- 2.0.0 (2026-06-23)
- Supported modalities
- RNA-seq, Affymetrix microarrays
- Statistical methods
- DESeq2, limma-voom
- Primary inputs
- Count/intensity matrices (TSV), sample sheets (CSV), contrast files (YAML), GTF annotations, Affymetrix CEL archives (.tar), gene sets (GMT)
- Primary outputs
- Differential statistical tables, Quarto HTML reports, exploratory plots, interactive Shiny application
- Container support
- Docker, Singularity
What you can use it for
- RNA-seq differential expression with DESeq2
- Test for differential transcript abundance across experimental groups using gene count matrices, sample annotations, and contrast files.
- RNA-seq linear modelling with limma-voom
- Perform differential abundance testing using limma-voom on scaled abundance matrices, such as length-scaled transcript counts.
- Affymetrix microarray analysis
- Process raw probe-level microarray intensities provided as an archive of CEL files to identify differentially expressed genes.
- Differential gene set enrichment
- Integrate pathway-level functional analysis by supplying GMT gene set files alongside RNA-seq count tables.
- Multi-configuration testing
- Run and compare multiple parameter combinations and differential methods in parallel via a dedicated multi-run paramsheet.
Run it here
The research agent runs nf-core/differentialabundance 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
- Cross-checks observation matrices, sample annotations, feature collections, and contrast specifications before executing differential steps.
- Provides predefined analysis profiles tailored to specific platforms and algorithms, including RNA-seq DESeq2, limma-voom, and Affymetrix arrays.
- Produces both a rendered Quarto HTML report and an exportable parameterised notebook archive for customised execution in RStudio.
- Builds an optional interactive Shiny application via shinyngs for interactive thresholding and exploratory mining.
Watch out for
- Limma-voom analyses require scaled count inputs, such as scaled TPM, because the workflow cannot use offset matrices directly.
- Differential method parameters cannot be manually overridden within an analysis profile and must instead be modified using dedicated profile combinations.
- Automated deployment of the interactive Shiny application requires an external shinyapps.io account and setup.
Questions
Can I run nf-core/differentialabundance without my own cluster?
The research agent runs nf-core/differentialabundance 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/differentialabundance under?
MIT, like every nf-core pipeline.
Which differential analysis methods are available?
The pipeline provides pre-configured profiles for DESeq2 (the default for RNA-seq) and limma-voom.
What input data formats are required for RNA-seq analysis?
RNA-seq workflows require a feature observation matrix (TSV), a sample sheet (CSV), a contrast specification file (YAML), and optionally a GTF annotation file.
Can the pipeline analyse raw microarray data?
Yes, Affymetrix microarray data can be processed by supplying raw CEL files collected inside a TAR archive alongside sample and contrast files.
How are the results visualised and reported?
Results are delivered as an HTML report generated from a Quarto notebook, a zip archive containing the RStudio-ready notebook and inputs, and an optional Shiny web application.
What count inputs should be supplied for limma-voom?
Because limma-voom does not support the offset matrix format, scaled counts derived from abundances (such as scaledTPM or lengthScaledTPM) should be used instead of raw counts.
How to cite it
Philip Ewels, Alexander Peltzer, Sven Fillinger, Harshil Patel, Johannes Alneberg, Andreas Wilm, Maxime Ulysse Garcia, Paolo Di Tommaso & Sven Nahnsen. The nf-core framework for community-curated bioinformatics pipelines. Nat Biotechnol. 2020 Feb 13. doi: 10.1038/s41587-020-0439-x.
Written from nf-core/differentialabundance's own README. Statements citing a figure that does not appear there are removed before publication. Licence and running details come from our registry.