Skip to content

Home

msmu logo

Python toolkit for modular and traceable LC-MS/MS proteomics analysis based on MuData

Start with Installation and the DDA-TMT tutorial. For development-only APIs, use the source checkout described in the installation guide.

Overview

msmu is an open-source Python package for modular and traceable post-DB search preprocessing and statistical analysis of bottom-up proteomics data.

It supports modules for every step of end-to-end processing—from search output parsing through hierarchical summarization, normalization, batch correction, statistical analysis, and visualization—implemented with commonly used analytical and statistical methods.

Central to msmu is the highly versatile and standardized MuData (and AnnData) as a unifying, provenance-aware data container for organizing and storing annotations and representations of multi-dimensional MS data and processing history.

This unique marriage between flexible processing pipeline and MuData empowers FAIR principle-aligned downstream analysis for biomarker discovery and systems biology.

msmu workflow from database search and sample metadata through processing, analysis, and visualization

Key features

  • Flexible data ingestion from Sage, DIA-NN, and other popular DB search tools
  • MuData/AnnData-compatible object structure for organizing multi-level MS data
  • SDRF-native sample metadata: attach a validated SDRF as the immutable source of truth and project it onto samples
  • Command provenance: supported processing calls record arguments, execution times, environment versions, lineage, and optional content hashes in the container; console output and dimension summaries are not stored
  • Protein inference: infer protein groups from peptide evidence using parsimony rule
  • Hierarchical summarization: PSM/precursor → peptide → protein/PTM rollup (median, mean, sum, median polish, and DirectLFQ)
  • Normalization: median centering, quantile normalization, etc.
  • Batch correction for discrete and continuous variations
  • Built-in QC: identification count, peptide length, charge, missed cleavage, intensity distribution, etc.
  • Statistical analysis: differential expression (limma moderated-t by default, or a label-permutation test), dimensionality reduction
  • PTM data support and stoichiometry adjustment with matched global dataset (if available)
  • Visualization: PCA, UMAP, volcano plots, heatmaps, and QC metrics

Supported DB search tools

Tool Supported inputs Repository Publication
Sage DDA (label-free or TMT) lazear/sage Lazear (2023)
DIA-NN DIA (label-free precursors) vdemichev/DiaNN Demichev et al. (2020)
MaxQuant DDA (label-free or TMT) JurgenCox/mqtools7, partial Cox and Mann (2008)
FragPipe DDA (label-free or TMT) Nesvilab/FragPipe Kong et al. (2017)
DelPi DIA (label-free precursors) bertis-informatics/delpi Park et al. (2026), preprint

Supported inputs describe the data that msmu can import from each tool. See Data ingestion for input files and reader options.

License

BSD 3-Clause License. See LICENSE for details.

Citation

If you use msmu in your research, please cite the following publication (preprint):

Choi, H.-W., Lee, B., Kang, U.-B., and Huh, S. msmu: a Python toolkit for modular and traceable LC-MS proteomics data analysis based on MuData. bioRxiv 2026.01.07.698308 (2026). https://doi.org/10.64898/2026.01.07.698308.

References

Lazear, M. R. Sage: an open-source tool for fast proteomics searching and quantification at scale. Journal of Proteome Research 22, 3652–3659 (2023). https://doi.org/10.1021/acs.jproteome.3c00486.

Demichev, V., Messner, C. B., Vernardis, S. I., Lilley, K. S., and Ralser, M. DIA-NN: neural networks and interference correction enable deep proteome coverage in high throughput. Nature Methods 17, 41–44 (2020). https://doi.org/10.1038/s41592-019-0638-x.

Cox, J. and Mann, M. MaxQuant enables high peptide identification rates, individualized p.p.b.-range mass accuracies and proteome-wide protein quantification. Nature Biotechnology 26, 1367–1372 (2008). https://doi.org/10.1038/nbt.1511.

Kong, A. T., Leprevost, F. V., Avtonomov, D. M., Mellacheruvu, D., and Nesvizhskii, A. I. MSFragger: ultrafast and comprehensive peptide identification in mass spectrometry–based proteomics. Nature Methods 14, 513–520 (2017). https://doi.org/10.1038/nmeth.4256.

Park, J., Kim, K., Kang, U.-B., and Kim, S. DelPi learns generalizable peptide–signal correspondence for mass spectrometry-based proteomics. bioRxiv 2026.01.06.697814 (2026), preprint. https://doi.org/10.64898/2026.01.06.697814.