The deep learning framework Hist2Prot infers single-cell spatial proteomics from histology to predict biomarkers and clinical outcomes.

Hist2Prot infers single-cell spatial proteomic profiles from hematoxylin and eosin histology, trained on ~51 million paired cells and validated across ~72 million additional cells.

• Why it matters: Routine clinical use of spatial proteomics is constrained by cost, technical complexity, and limited tissue coverage, leaving a gap for scalable, tissue-wide protein mapping.
• What they did: The deep learning framework was trained on a same-section G4X dataset and evaluated on internal tests, independent same-platform cohorts, and external cross-platform datasets using concordance with measured protein profiles.
• The result: In seven patient cohorts totaling 1,774 patients, virtual spatial proteomics improved biomarker, prognostic, and immunotherapy response prediction beyond hematoxylin and eosin morphology alone.

AI summary of the abstract · may contain errors · not medical advice · read the paper.