An open-source ecosystem for large-scale quantitative proteomics. Cloud-ready Nextflow pipelines, standalone Python libraries, and ML-powered tools — process DDA and DIA data with dedicated workflows, or use the standalone tools with any search engine.
# DDA proteomics pipeline (LFQ & TMT/iTRAQ) nextflow run bigbio/quantms \ -profile docker \ --input samplesheet.csv \ --database uniprot.fasta \ --outdir results/ # Runs on laptops, HPC, or AWS/GCP/Azure # DDA-LFQ, DDA-ISO (TMT/iTRAQ) workflows
# DIA proteomics pipeline (powered by DIA-NN) nextflow run bigbio/quantmsdiann \ -profile docker \ --input samplesheet.csv \ --database uniprot.fasta \ --outdir results/ # Dedicated DIA workflow with DIA-NN engine # Library-free or spectral library modes
# Convert search engine outputs to unified format qpx convert \ --tool diann \ --report diann_report.tsv \ --sdrf experiment.sdrf.tsv # Supported: DIA-NN, MaxQuant, FragPipe, # Spectronaut, quantms, Sage, MSFragger # Export to Parquet or AnnData (.h5ad)
# Protein quantification from any search engine # Works with quantms, quantmsdiann, MaxQuant, FragPipe mokume features2proteins \ -p features.parquet \ -o proteins.csv \ -s experiment.sdrf.tsv \ --quant-method maxlfq # TMT with IRS + differential expression mokume features2proteins \ -p features.parquet \ --quant-method median \ --irs --de
# Interactive QC reports for any experiment multiqc results/ \ --module pmultiqc # Generate mokume QC + comparison reports mokume features2proteins \ -p features.parquet \ -s experiment.sdrf.tsv \ --report qc_report.html # HTML reports: PCA, heatmaps, CVs, volcanos
A Nextflow-orchestrated workflow where specialized tools handle each step.
Two dedicated Nextflow pipelines: quantms for DDA (LFQ & TMT/iTRAQ) and quantmsdiann for DIA — but the downstream tools work independently too. Use mokume to quantify proteins from any search engine output. Use pmultiqc for QC on any experiment.
Browse reanalyzed proteomics datasets and explore results interactively.
The quantms.org data portal hosts thousands of public proteomics datasets reanalyzed from ProteomeXchange repositories (PRIDE, MassIVE, jPOST). Browse absolute and differential expression across tissues, diseases, and organisms. All ecosystem tools also work locally for your own data analysis.
Deep learning models trained on 500M+ spectra, powering the next generation of proteomics analysis.
The quantms ecosystem reanalyzes thousands of public proteomics experiments, producing standardized datasets. These datasets train deep learning models that improve every step of the pipeline — from spectrum prediction to de novo sequencing.
Built on open formats for reproducible, metadata-rich proteomics.
Every experiment in the ecosystem is described with SDRF (Sample and Data Relationship Format), ensuring complete metadata annotation. The qpx library bridges the gap between search engines — converting outputs from quantms, quantmsdiann, DIA-NN, FragPipe, MaxQuant, and Spectronaut into a unified format.
Convert outputs from any search engine into a unified Parquet format. Export to AnnData for integration with scverse and other analysis ecosystems.
qpx documentation →Standardized experimental design: organisms, tissues, labels, conditions. Used by all quantms tools for automatic sample grouping.
Convert from DIA-NN, FragPipe, MaxQuant, Spectronaut, quantms. Unified Parquet output with PSM-level data.
Export protein matrices to .h5ad for integration with scanpy, scverse, and the broader Python data science ecosystem.
mokume — the protein quantification engine at the heart of quantms.
A comprehensive Python library for protein quantification from peptide-level data. Works standalone with output from any search engine — not just quantms. Supports 6+ quantification methods, normalization, IRS for multi-plex TMT, batch correction, and differential expression.
# LFQ quantification mokume features2proteins \ -p features.parquet \ -o proteins.csv \ -s experiment.sdrf.tsv \ --quant-method maxlfq # TMT + IRS + DE analysis mokume features2proteins \ -p features.parquet \ --quant-method median \ --irs --de \ --de-contrasts "NASH-HL"
from mokume.pipeline import ( QuantificationPipeline, PipelineConfig ) from mokume.pipeline.config import ( InputConfig, QuantificationConfig ) config = PipelineConfig( input=InputConfig( parquet="features.parquet", sdrf="experiment.sdrf.tsv", ), quantification=QuantificationConfig( method="maxlfq", ), ) pipeline = QuantificationPipeline(config) proteins = pipeline.run()
// Run quantms pipeline nextflow run bigbio/quantms \ -profile docker \ --input samplesheet.csv \ --database uniprot.fasta \ --outdir results/ // Cloud-ready: AWS, GCP, Azure // Nature Methods 2024
Publication-ready QC reports and interactive visualizations.
pmultiqc generates interactive HTML reports with comprehensive QC metrics for every experiment. mokume adds differential expression reports, interactive plots, and workflow comparison dashboards.
Eight specialized tools, one unified platform.
DIA proteomics powered by DIA-NN. Library-free and spectral library modes.
Protein quantification, normalization, batch correction, DE. Successor to ibaqpy.
SDRF validation, annotation, and conversion for proteomics metadata.
SDRF metadata, open file formats, and reproducible workflows ensure every experiment is fully documented and shareable.
Process 50,000+ datasets at cloud scale. Nextflow orchestration with Docker/Singularity runs on laptops, HPC clusters, or AWS/GCP/Azure. Python libraries handle local analysis just as well.
Not locked in. Use mokume and pmultiqc with data from quantms, quantmsdiann, MaxQuant, FragPipe, Spectronaut, or DIA-NN. Export to AnnData for the scverse ecosystem.
Interactive QC reports, volcano plots, PCA, heatmaps, and differential expression tables — ready for your next manuscript.
Join 20+ contributors from EMBL-EBI, OpenMS, and research institutions worldwide.
Contribute on GitHub