# UNI2-h

2 related articles

Subscribe to updates

bioRxiv (Cold Spring Harbor Laboratory) Oct 25, 2026
H&E slides skip spatial transcriptomics: ALHENA distills cell and neighborhood similarity, reaching 0.341 patient-level signature correlation
ALHENA distills spatial transcriptomic information into a routine H&E encoder that reads only images at inference. Rather than regressing individual genes, it trains the H&E representation to match the sample-to-sample similarity structure of two frozen transcriptomic models at cell and neighborhood scales. Across 110 cancer-type and signature combinations in TCGA, mean Spearman correlation was 0.341 versus 0.161 for unadapted UNI2-h; on the Thunder pathology benchmark the median change was −0.004.
ALHENA distills spatial transcriptomic information into a routine H&E encoder that reads only images at inference. Rather than regressing individual genes, it trains the H&E representation to match the sample-to-sample similarity structure of two frozen transcriptomic models at cell and neighborhood scales. Across 110 cancer-type and signature combinations in TCGA, mean Spearman correlation was 0.341 versus 0.161 for unadapted UNI2-h; on the Thunder pathology benchmark the median change was −0.004.
Inserm Université Paris-Saclay
Félicie Giraud-Sauveur（First author, Corresponding author）Paul-Henry Cournède（Corresponding author）Stergios Christodoulidis（Corresponding author）
ALHENA Multi-scale contrastive distillation Spatial transcriptomics Gene expression prediction from histology Cell Type Annotation
arXiv Oct 30, 2026
One H&E slide, three transcriptomes: a multi-scale graph network beats DeepSpot2Cell in cell-level PCC and AUROC
Spatial and single-cell transcriptomics reveal where genes are expressed in tissue, but each slide is expensive and scarce, while routine H&E whole-slide images are abundant. A multi-scale graph neural network trains on TCGA bulk data, Visium spatial data and Xenium single-cell data together, using consistency losses to keep predictions aligned across the three scales, and predicts bulk, spatial and single-cell expression directly from slide images. It performs best at slide level and moderately at patch level, with single-cell prediction the hardest; on Xenium its cell-level PCC and AUROC exceed DeepSpot2Cell, while DeepSpot2Cell leads at patch level.
Spatial and single-cell transcriptomics reveal where genes are expressed in tissue, but each slide is expensive and scarce, while routine H&E whole-slide images are abundant. A multi-scale graph neural network trains on TCGA bulk data, Visium spatial data and Xenium single-cell data together, using consistency losses to keep predictions aligned across the three scales, and predicts bulk, spatial and single-cell expression directly from slide images. It performs best at slide level and moderately at patch level, with single-cell prediction the hardest; on Xenium its cell-level PCC and AUROC exceed DeepSpot2Cell, while DeepSpot2Cell leads at patch level.
Miguel Andrade（First author）
Graph Neural Network Spatial transcriptomics Gene expression prediction from whole-slide images UNI2-h Survival analysis

You’re all caught up

H&E slides skip spatial transcriptomics: ALHENA distills cell and neighborhood similarity, reaching 0.341 patient-level signature correlation

One H&E slide, three transcriptomes: a multi-scale graph network beats DeepSpot2Cell in cell-level PCC and AUROC