nf-core pipeline · dda, immunopeptidomics, mass-spectrometry
nf-core/mhcquant
Identifies and quantifies MHC-eluted peptides from data-dependent acquisition mass spectrometry datasets.
What nf-core/mhcquant is
nf-core/mhcquant is a bioinformatics pipeline designed for the analysis of data-dependent acquisition (DDA) immunopeptidomics data. Immunopeptidomics focuses on the isolation and identification of peptides presented on major histocompatibility complex (MHC) molecules, which guide T-cell immunosurveillance and inform the development of targeted immunotherapies.
Built using the OpenMS framework for computational mass spectrometry and implemented in Nextflow DSL2, the pipeline standardises peptide identification and quantification across heterogeneous compute environments. It was originally developed by Leon Bichmann and rewritten by Marissa Dubbelaar, with further developments led by Jonas Scheid and Steffen Lemke.
The workflow searches mass spectra against a target protein database, refines identifications using predicted peptide properties alongside machine-learning rescoring, and exports false discovery rate-filtered peptide lists. It also offers workflows for label-free quantification, spectrum library construction, and epitope consensus prediction.
What the pipeline does
Input mass spectrometry data (such as Thermo raw, Bruker tdf, or mzML files) are prepared alongside a FASTA protein database, from which a reversed decoy sequence database is constructed. Peptide-spectrum matches (PSMs) are identified using the Comet search engine with default settings targeted at MHC class I peptides fragmented by higher-energy collisional dissociation (HCD). Protein references and target-decoy assignments are maintained via PeptideIndexer, and identifications sharing sample and condition descriptors are merged.
Identifications then pass through the RESCORE subworkflow, where MS²Rescore predicts peptide retention times and MS2 fragment intensities. Extracted PSM features are evaluated by Percolator or Mokapot to estimate false discovery rates (FDR) under either local (per sample-condition) or global (dataset-wide) FDR control modes. When quantification is enabled, the pipeline executes retention time alignment with MapAlignerIdentification, detects MS1 features with FeatureFinderIdentification, links features across runs, and resolves identification conflicts.
Key facts
- Workflow framework
- Nextflow DSL2
- Core computational engine
- OpenMS
- Sequence search engine
- Comet
- Rescoring tools
- MS²Rescore, Percolator, Mokapot
- Input spectrum formats
- raw, RAW, mzML, mzML.gz, d, d.tar.gz, d.zip
- Sequence input format
- FASTA
- Default analysis target
- MHC class I peptides with HCD fragmentation
- FDR control modes
- Local FDR (per sample-condition) and Global FDR (dataset-wide)
What you can use it for
- MHC Class I immunopeptide identification
- Search eluted ligand mass spectra against reference proteomes to identify HLA-presented peptides using machine-learning rescoring.
- Label-free immunopeptide quantification
- Align retention times and quantify peptide feature intensities across different experimental conditions or disease states.
- Spectral library generation for DIA
- Generate empirical spectrum libraries from DDA experiments for downstream data-independent acquisition mass spectrometry searches.
- Consensus epitope derivation
- Compute consensus epitope representations across detected immunopeptides using integrated epicore processing.
Run it here
The research agent runs nf-core/mhcquant 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
- Incorporates retention time and fragment intensity predictions from MS²Rescore to boost Percolator and Mokapot rescoring performance.
- Natively supports native vendor formats from Thermo and Bruker in addition to open mzML formats.
- Supports both condition-specific local FDR and multi-sample global FDR control depending on study design.
- Provides identification-based alignment and feature linking for label-free quantification across multiple runs.
Watch out for
- Configured by default exclusively for MHC class I peptides analysed with HCD fragmentation.
- Label-free quantification, spectral library creation, and epicore consensus epitope generation are disabled by default and require explicit command-line flags.
- Pipeline parameters cannot be supplied via custom Nextflow configuration files and must be specified via CLI flags or a parameter file.
Questions
Can I run nf-core/mhcquant without my own cluster?
The research agent runs nf-core/mhcquant 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/mhcquant under?
MIT, like every nf-core pipeline.
Which mass spectrometer raw file formats are supported?
The pipeline accepts Thermo raw files (.raw, .RAW), Bruker files (.d, .d.tar.gz, .d.zip), and mzML files (.mzML, .mzML.gz).
How does the pipeline handle false discovery rate calculation?
It applies a competitive target-decoy approach rescored by Percolator or Mokapot. FDR filtering can be applied locally within sample-condition groups (default) or globally across the complete dataset using the --global_fdr option.
How is label-free quantification enabled?
Quantification is activated using the --quantify flag, which initiates retention time alignment, MS1 feature detection, cross-run feature grouping, and conflict resolution.
Can the pipeline produce libraries for DIA analyses?
Yes. Passing the --generate_speclib flag produces spectrum libraries suitable for DIA searches, yielding either individual sample libraries or a single global library when combined with global FDR mode.
How to cite it
Jonas Scheid, Steffen Lemke, Naomi Hoenisch-Gravel, Anna Dengler, Timo Sachsenberg, Arthur Declerq, Ralf Gabriels, Jens Bauer, Marcel Wacker, Leon Bichmann, Lennart Martens, Marissa L. Dubbelaar, Sven Nahnsen & Juliane S. Walz. MHCquant2 refines immunopeptidomics tumor antigen discovery. Genome Biology 2025 26 (1), 290. doi: 10.1186/s13059-025-03763-8
Written from nf-core/mhcquant's own README. Statements citing a figure that does not appear there are removed before publication. Licence and running details come from our registry.