Reproduce “Predicting Single-Cell Spatial Gene Expression from H&E using Pathology Foundation Models” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Ashley P. Tsang, Evan Liu, Conrad Foo, Bob Chen, Hassaan Maan, Donny Chan, Elisa Penna, Veronica Ibarra Lopez and 5 more
The authors adapt tile-trained pathology foundation models to predict single-cell gene expression from colon H&E images, selecting and pooling image features within each target cell and training on Xenium counts. Across five block-level test folds on the 5K panel, the model achieved per-gene PCC of 0.174 ± 0.014, outperforming SpatialEx; performance varies by gene and does not establish broader generalization.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.