nf-core pipeline · differential-abundance, micobiome
nf-core/daamicrobiome
Microbiome differential abundance pipeline combining multiple tools via k-intersection or simulation consensus.
What nf-core/daamicrobiome is
nf-core/daamicrobiome is a bioinformatics pipeline designed for microbiome differential abundance analysis, originally developed by Martina Cardinali and Toni Gabaldón. Differential abundance analysis aims to identify operational taxonomic units (OTUs) or amplicon sequence variants (ASVs) that differ significantly in abundance across experimental groups or clinical conditions. Because different statistical testing methods often produce divergent findings on microbiome count data, the pipeline executes several differential abundance tools in parallel and consolidates their findings.
The pipeline ingests a phyloseq object saved in R RDS format, which must contain an OTU or ASV count matrix, sample metadata including an experimental group column, and taxonomic classifications. Users can run their analysis through two alternative consensus strategies depending on whether they wish to benchmark tool accuracy against simulated controls.
What the pipeline does
The workflow supports two operational routes: Path A (k-intersection consensus) and Path B (simulation-based weighted consensus, which is the default). In Path A, the pipeline runs differential abundance tools directly on experimental data, outputting any taxa identified as differentially abundant by at least k distinct tools.
In Path B, control or healthy samples are isolated from the input phyloseq object or supplied as a pre-filtered control object. The MIDASim package then generates synthetic benchmark datasets containing known differentially abundant taxa based on these controls. The enabled differential abundance tools—which can include ADAPT, corncob, LinDA, LOCOM, MaAsLin2, and metagenomeSeq—are evaluated on the simulated datasets to assess sensitivity, specificity, and false discovery rate. These performance metrics determine tool-specific weights, which are subsequently applied to aggregate the real-data differential abundance findings into a weighted consensus.
Key facts
- Input format
- phyloseq RDS object (.RDS)
- Included tools
- ADAPT, corncob, LinDA, LOCOM, MaAsLin2, metagenomeSeq
- Simulation tool
- MIDASim
- Consensus strategies
- k-intersection (Path A) and simulation-based weighted consensus (Path B)
- Pipeline authors
- Martina Cardinali and Toni Gabaldón
- Workflow framework
- Nextflow (nf-core)
What you can use it for
- Weighted consensus differential abundance
- Benchmarking multiple differential abundance tools on MIDASim-simulated control data to weight and merge real-data findings according to tool accuracy.
- Intersection-based multi-tool analysis
- Filtering differential taxa from microbiome profiles by requiring agreement across at least k statistical packages without running simulations.
- Method concordance evaluation
- Executing ADAPT, corncob, LinDA, LOCOM, MaAsLin2, and metagenomeSeq simultaneously across the same experimental dataset to inspect agreement.
Run it here
The research agent runs nf-core/daamicrobiome 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
- Executes up to six established microbiome differential abundance methods in parallel within a single pipeline run.
- Derives objective, dataset-specific tool weights by simulating ground-truth data from healthy or control samples using MIDASim.
- Offers both an unweighted threshold-based consensus (k-intersection) and a simulation-trained weighted consensus approach.
Watch out for
- Input data must be structured specifically as an R phyloseq RDS object containing count data, taxonomy tables, and metadata.
- The default weighted consensus mode requires designated control or healthy samples to train the MIDASim simulation models.
Questions
Can I run nf-core/daamicrobiome without my own cluster?
The research agent runs nf-core/daamicrobiome 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/daamicrobiome under?
MIT, like every nf-core pipeline.
What input data does the pipeline require?
The pipeline requires a phyloseq RDS object (.RDS) containing an OTU/ASV count table, sample metadata with a condition or group column, and a taxonomy table.
Which differential abundance packages are available?
The pipeline includes ADAPT, corncob, LinDA, LOCOM, MaAsLin2, and metagenomeSeq.
How do Path A and Path B differ?
Path A runs differential abundance tools directly on experimental samples and reports taxa called significant by at least k tools. Path B uses control samples and MIDASim to generate simulated data, scores each tool's sensitivity, specificity, and false discovery rate on those simulations, and applies the resulting weights to the real data.
How are control samples specified for simulations?
The pipeline can subset control or healthy samples directly from the input phyloseq object based on a condition parameter and baseline level, or it can accept a separate control-only phyloseq object.
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/daamicrobiome's own README. Statements citing a figure that does not appear there are removed before publication. Licence and running details come from our registry.