Skip to main content

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

  • Correspondence
  • Published:

Unpaired data as a first-order challenge in single-cell and spatial proteomics

Nature Biotechnology (2026)Cite this article

Subjects

Single-cell and spatial proteomics are rapidly maturing, with mass spectrometry and imaging-based technologies now capable of quantifying tens to thousands of proteins across cells and tissues. Yet, as measurement resolution, depth and throughput improve, progress is increasingly constrained not by experimental sensitivity but by the ability to integrate proteomic data across experiments, modalities and spatial contexts. In practice, most proteomic measurements are unpaired to transcriptomic or epigenomic references: proteins are measured in different cells, adjacent tissue sections or separate experiments. These partial, misaligned, and heterogeneous observations represent the dominant data regime in current studies, but they remain poorly supported by existing analytical frameworks1.

Much of the single-cell analysis ecosystem was developed under assumptions that do not hold for proteomics. Transcriptomics workflows typically rely on dense, paired measurements, with one-to-one correspondences between cells across conditions or modalities. In contrast, proteomics datasets, particularly those generated by mass spectrometry or spatial imaging, are often sparse, incomplete, and non-overlapping by design. Spatial proteomics adds further complexity: protein maps may need to be registered across tissue sections, platforms or resolutions, frequently without a shared cellular index. Treating these scenarios as exceptions rather than defaults leads to brittle analyses, limits cross-study comparison, and risks obscuring biological signal behind technical mismatches2,3. Taken together, these constraints indicate that pairing should no longer be treated as the default assumption for multimodal proteomics analyses.

This is a preview of subscription content, access via your institution

Access options

Buy this article

  • Purchase on SpringerLink
  • Instant access to the full article PDF.

USD 39.95

Prices may be subject to local taxes which are calculated during checkout

References

  1. Bennett, H. M., Stephenson, W., Rose, C. M. & Darmanis, S. Nat. Methods 20, 363–374 (2023).

    Article  PubMed  CAS  Google Scholar 

  2. Boekweg, H. & Payne, S. H. Mol. Cell. Proteomics 22, 100518 (2023).

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  3. Dai, C. et al. Nat. Methods 21, 1603–1607 (2024).

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  4. Virshup, I. et al. Nat. Biotechnol. 41, 604–606 (2023).

    Article  PubMed  CAS  Google Scholar 

  5. Furtwängler, B. et al. Science 390, eadr8785 (2025).

    Article  PubMed  Google Scholar 

  6. Cao, Z.-J. & Gao, G. Nat. Biotechnol. 40, 1458–1466 (2022).

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  7. Nat. Methods 21, 2195–2196 (2024).

  8. Marconato, L. et al. Nat. Methods 22, 58–62 (2025).

    Article  PubMed  CAS  Google Scholar 

Download references

Acknowledgements

We thank all members of the scverse community for their contribution to this manuscript. This study was supported by Veterans Affairs Merit grant BX004189 (to P.R.) and by the US National Institutes of Health: R01MH125246 (to P.R.), R01AG067025 (to P.R.), R01AG082185 (to P.R.), R01AG065582 (to P.R.), R01AG050986 (to P.R.), R01AG095776 (to P.R.), U01MH116442 (to P.R.) and U24AG087563 (to P.R.). Research in the lab of E.M.S. is supported by the Lundbeck Foundation (R413-2022-869), the Independent Research Fund Denmark (2067-00053B) and the Velux Foundation (00053026).

Author information

Authors and Affiliations

  1. Genentech, South San Francisco, CA, USA

    Mikaela Koutrouli, Carolina Leote, Christopher M. Rose & John C. Marioni

  2. Department of Molecular Biology and Digital Science Center (DiSC), University of Innsbruck, Innsbruck, Austria

    Francesca Finotello & Felix Petschko

  3. Department of Biotechnology and Biomedicine, Technical University of Denmark, Kongens Lyngby, Denmark

    Erwin M. Schoof, Pedro Aragon-Fernandez & Valdemaras Petrosius

  4. Center for Disease Neurogenomics, Icahn School of Medicine at Mount Sinai, New York, NY, USA

    Panos Roussos, Prashant FNU & Gabriel E. Hoffman

  5. European Molecular Biology Laboratory (EMBL), Heidelberg, Germany

    Wouter-Michiel Vierdag, Matthias Meyer-Bender, Harald Vöhringer, Luca Marconato & Wolfgang Huber

  6. Department of Proteomics and Signal Transduction, Max Planck Institute of Biochemistry, Martinsried, Germany

    Lucas Diedrich, Vincenth Brennsteiner & Matthias Mann

  7. Department of Clinical Medicine, Aarhus University, Aarhus, Denmark

    Malte Kuehl & Victor G. Puelles

  8. Department of Pathology, Aarhus University Hospital, Aarhus, Denmark

    Malte Kuehl & Victor G. Puelles

  9. Max Delbrück Center for Molecular Medicine in the Helmholtz Association (MDC), Berlin, Germany

    Jose Nimo & Fabian Coscia

  10. Charité – Universitätsmedizin Berlin, Berlin, Germany

    Jose Nimo & Fabian Coscia

  11. Maastricht MultiModal Molecular Imaging (M4I) Institute, Division of Imaging Mass Spectrometry, Maastricht University, Maastricht, the Netherlands

    Theodoros Visvikis, Ron M. A. Heeren & Eva Cuypers

  12. Division of Computational Genomics and System Genetics, German Cancer Research Center, Heidelberg, Germany

    Luca Marconato

  13. Institute of Computational Biology, Helmholtz Center Munich, Munich, Germany

    Sergio Marco Salas & Fabian J. Theis

  14. Department of Information Technology, Uppsala University, Uppsala, Sweden

    Sergio Marco Salas

  15. napari https://napari.org/stable/

    Lorenzo Gaifas

  16. Genome Biology Unit, European Molecular Biology Laboratory (EMBL), Heidelberg, Germany

    Sinem K. Saka

  17. TUM School of Computation, Information and Technology, Technical University of Munich, Munich, Germany

    Fabian J. Theis

  18. TUM School of Life Sciences, Technical University of Munich, Munich, Germany

    Fabian J. Theis

Authors
  1. Mikaela Koutrouli
    View author publications

    Search author on:PubMed Google Scholar

  2. Francesca Finotello
    View author publications

    Search author on:PubMed Google Scholar

  3. Erwin M. Schoof
    View author publications

    Search author on:PubMed Google Scholar

  4. Panos Roussos
    View author publications

    Search author on:PubMed Google Scholar

Consortia

on behalf of The scverse Proteomics Consortium

  • Mikaela Koutrouli
  • , Francesca Finotello
  • , Erwin M. Schoof
  • , Panos Roussos
  • , Wouter-Michiel Vierdag
  • , Lucas Diedrich
  • , Matthias Meyer-Bender
  • , Malte Kuehl
  • , Jose Nimo
  • , Harald Vöhringer
  • , Theodoros Visvikis
  • , Vincenth Brennsteiner
  • , Luca Marconato
  • , Pedro Aragon-Fernandez
  • , Felix Petschko
  • , Sergio Marco Salas
  • , Carolina Leote
  • , Christopher M. Rose
  • , Prashant FNU
  • , Gabriel E. Hoffman
  • , Ron M. A. Heeren
  • , Lorenzo Gaifas
  • , Valdemaras Petrosius
  • , Victor G. Puelles
  • , Eva Cuypers
  • , Matthias Mann
  • , Fabian Coscia
  • , Sinem K. Saka
  • , Wolfgang Huber
  • , John C. Marioni
  •  & Fabian J. Theis

Contributions

M. Koutrouli and P.R. wrote the manuscript. All authors contributed to the content and read and approved the final version of the manuscript.

Corresponding authors

Correspondence to Mikaela Koutrouli or Panos Roussos.

Ethics declarations

Competing interests

F.J.T. consults for Immunai, CytoReason, Genbio, Valinor Industries, Bioturing and Phylo Inc., and has ownership interest in RN.AI Therapeutics, Dermagnostix and Cellarity. The remaining authors declare no competing interests.

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Koutrouli, M., Finotello, F., Schoof, E.M. et al. Unpaired data as a first-order challenge in single-cell and spatial proteomics. Nat Biotechnol (2026). https://doi.org/10.1038/s41587-026-03074-8

Download citation

  • Published:

  • Version of record:

  • DOI: https://doi.org/10.1038/s41587-026-03074-8

Search

Advanced search

Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing