🙏 Thanks to all collaborators at the DKFZ, UKHD and Heidelberg University, and UKE Institute of Pathology, to the patients, and funders.
@saezlab.bsky.social
@dkfz.bsky.social
@dkfz-iic.bsky.social
@tanevski.bsky.social
@struxa.bsky.social
@hlsalliance.bsky.social
@erc.europa.eu
@helmholtz.de
5/5

Our analysis reveals distinct recurrent CRC trajectories:
🤝 Tumor-CAF metabolic cooperation with immune shielding
🔥 Glycolytic shift with vascular remodeling and myeloid enrichment
Same disease, different mechanisms, suggesting different treatment responses.
4/5

🧩 To identify which changes co-occur in patients, we built MuVIcell, an open-source Python package that applies the MuVI factor analysis framework to model all spatial, metabolic, morphological, and compositional features per cell type.
Try it yourself: github.com/HartmannLab/...
3/5

⏳ What shifts with pT stage?
Early lesions show high proliferation and fatty acid oxidation capacity. Later, glycolysis rises, stroma and vasculature reorganize, and lymphocytes give way to myeloid cells. Spatial architecture predicts stage better than composition or metabolic state.
2/5