Technology Advances
On the technology front, we introduced several transformative methods. Our SPEC (Solid-Phase Extraction Capture) workflow enables proteomics sample preparation in nanoliter sample volumes resulting in unprecedented sensitivity. It enables the nanoPhos method which brings phosphoproteomics—traditionally requiring substantial material—to single-cell resolution in spatial contexts.
AlphaDIA, published in Nature Biotechnology, pioneers end to end deep learning and applies transfer learning to adjust to new conditions and PTM types, further extending our Bioinformatics toolbox. In the Rust rewrite it has become blazingly fast. We also developed ADAPT-MS, an adaptive machine learning framework that enables clinical decision-making directly from discovery proteomics data without requiring fixed biomarker panels—a paradigm shift for clinical proteomics.
Proteomics Methods
Nanoliter-volume workflow for fast, robust and ultrasensitive proteomics enabling unprecedented sensitivity from minimal sample input.
Simplified perchloric acid workflow enabling deep plasma proteomics at population scale with >10,000 samples per day throughput.
Ultra-sensitive cell-type resolved spatial phosphoproteomics enabling signaling analysis at single-cell resolution in tissues.
Accessible step-by-step protocol for high-sensitivity proteomics using parallel accumulation-serial fragmentation.
Novel hybrid high-speed mass spectrometer enables rapid translation from biomarker candidates to targeted clinical tests using 15N-labeled proteins.
Systematic evaluation of plasma proteomics workflows reveals susceptibility to cellular contamination and provides framework for quality control.
Bioinformatics and Software
End-to-end transfer learning enabling feature-free proteomics with dramatically improved identification rates.
Adaptive diagnostic framework for clinical proteomics enabling personalized testing from plasma and CSF.
Software package for generation, storage and application of single-cell image datasets enabling multimodal analysis.
Novel computational approach to infer proteoform regulation from bottom-up proteomics data.
Multimodal AI agents for capturing and sharing tacit laboratory practice through video, speech, and text analysis.
High-Throughput Applications
Our Deep Visual Proteomics (DVP) platform continued to yield biological insights with clinical relevance. In collaboration with clinical partners, we revealed why some patients with alpha-1 antitrypsin deficiency develop severe liver disease while others remain healthy.
Disease Biology and Spatial Proteomics
Deep Visual Proteomics reveals why some patients with the hereditary liver disease remain healthy while others develop severe pathology.
Single cell spatial proteomics maps human hepatocyte zonation patterns and their vulnerability to fibrosis.
DVP reveals in vivo-like phenotype upon xenotransplantation of human colon organoids.
Cell type-resolved proteomics and phosphoproteomics decode adult murine pancreatic islet cell diversity.
DNA replication stress identified as a hallmark of signet ring cell carcinoma through DVP.
Deciphering functional tumor-immune crosstalk through highly multiplexed imaging and deep visual proteomics.
We continue to work on bringing large-scale proteomics studies to the patients and into the clinic. Beyond the ADAPT-MS framework, we also established a new framework for studying rare diseases. Our ontology-guided clustering approach enables meaningful proteomic analysis even when individual conditions affect only a handful of patients—opening the door to systematic study of the thousands of rare pediatric disorders that collectively affect millions of children worldwide.
Clinical and Population Proteomics
Comprehensive proteomic profiling of 2,147 children revealing associations with age, sex, puberty, BMI, and genetics.
Framework integrating clinical ontologies with proteomics enables analysis of 394 rare pediatric conditions from 1,140 patients.
Major Collaborations
We were involved in multiple successfull collaborations, contributing with our expertise in MS-technologies and bioinformatics. In particular, our long-standing collaboration with Ernst Lengyel at the University of Chicago produced multiple high-impact papers on ovarian cancer, including the identification of NNMT as a therapeutic target in cancer-associated fibroblasts (Nature) and comprehensive spatial mapping of serous tubal intraepithelial carcinomas (STICs), revealing that ovarian cancer precursors are far more common than previously recognized.
Ovarian Cancer
Spatial proteo-transcriptomics reveals molecular landscape of borderline ovarian tumors. Identified 16 drug targets; combination therapy achieved significant tumor reduction.
Integration of cell-type resolved spatial proteomics and transcriptomics reveals novel mechanisms including SUMOylation activation and ATR signaling.
NNMT inhibitor reduces tumor burden by reprogramming cancer-associated fibroblasts and restoring immune checkpoint blockade efficacy.
Spatially resolved multi-omics framework identifies 99 STICs and precursors in cancer-free organ donors, revealing high incidence of ovarian cancer precursors.
E3 ligases – Brenda Schulman (MPI of Biochemistry)
Clinical Proteomics - Nicolai Wewer Albrechtsen (Copenhagen)
This report highlights the potential of proteomics to inform Machine Learning-assisted diagnostics.
Other Key Collaborations
A fin-loop-like structure in GPX4 underlies neuroprotection from ferroptosis. Proteomics revealed Alzheimer's-like signatures.
Targeting specific LRRK2 kinase substrates rescues colitis severity caused by Crohn's disease-linked variant.
Clinical Impact Highlight
JAK Inhibitors for Toxic Epidermal Necrolysis
Our 2024 Nature publication (Nordmann TM et al.) on spatial proteomics identifying JAK inhibitors as treatment for toxic epidermal necrolysis (TEN) – a lethal skin disease – has had immediate clinical impact in 2025. Off-label use of JAK inhibitors in the first patients worldwide led to complete recovery.
“To our knowledge, this is the first time a spatial omics technology has made an immediate and tangible impact in the clinic, by identifying a treatment that has already changed people’s lives for the good.”
– Matthias Mann
Training and Recognition
Our commitment to training the next generation of scientists was reflected in the successful graduation of four PhD students this year, each of whom made significant contributions to our research program and are now launching careers in academia and industry.
The laboratory’s contributions to science were recognized with Mattias Mann’s election to the United States National Academy of Sciences, one of the highest honors in scientific research.
Publication Summary by Journal
| Journal | Count | Topics |
|---|---|---|
| Nature | 2 | Liver disease (DVP), Review |
| Nature Biotechnology | 1 | AlphaDIA transfer learning |
| Nature Genetics | 1 | Pediatric plasma proteome |
| Nature Protocols | 1 | PASEF workflow |
| Nature Communications | 1 | ADAPT-MS (in press) |
| Cancer Cell | 1 | Ovarian cancer progression (w/ Lengyel) |
| Cell Systems | 1 | Colon organoid DVP |
| EMBO Mol Med | 2 | Rare pediatric disorders, Pre-analytical bias |
| Molecular Cell | 1 | Tumor-immune crosstalk |
| Mol Cell Proteomics | 2 | PCA-N plasma, Stellar MS |
| Commun Biol | 1 | Pancreatic islet cells |
| NPJ Precision Oncol | 1 | Signet ring cell carcinoma |
| Mol Syst Biol | 1 | AI laboratory agents |