# Spatial Proteomics

Strand AI ran its virtual staining model Lattice over the TCGA colon adenocarcinoma cohort to produce predicted spatial protein maps for 442 diagnostic H&E slides and released them publicly. The resource covers seven protein markers, CD3e, CD8, CD20, CD68, Ki67, CD45 and PanCK, plus DAPI, and lets users compare original tissue with predicted channels over the same field of view without signing in, or access it through Hugging Face or a Python SDK. The release states that these predictions carry a prognostic signal in this cohort, with details to follow.

Strand AI ran its virtual staining model Lattice over the TCGA colon adenocarcinoma cohort to produce predicted spatial protein maps for 442 diagnostic H&E slides and released them publicly. The resource covers seven protein markers, CD3e, CD8, CD20, CD68, Ki67, CD45 and PanCK, plus DAPI, and lets users compare original tissue with predicted channels over the same field of view without signing in, or access it through Hugging Face or a Python SDK. The release states that these predictions carry a prognostic signal in this cohort, with details to follow.

The authors present SpaCEy (Spatial Clinical Explainability), an explainable graph neural network that builds each tissue sample into a spatial graph via Delaunay triangulation, with nodes as single-cell protein-marker abundances and no predefined cell-type or anatomical-region inputs, learns predictive embeddings with a GNN, and uses a GNNExplainer-style explainer to output edge masks that are aggregated over k-hop neighbourhoods into node importance, thereby localising contiguous outcome-associated spatial regions and key proteins; it predicted progression in a 416-patient lung adenocarcinoma cohort (accuracy 0.68, F1 0.68, AUC 0.62, versus Ali et al. 0.61/0.59/0.57 and SPACE-GM 0.55/0.52/0.

The authors present SpaCEy (Spatial Clinical Explainability), an explainable graph neural network that builds each tissue sample into a spatial graph via Delaunay triangulation, with nodes as single-cell protein-marker abundances and no predefined cell-type or anatomical-region inputs, learns predictive embeddings with a GNN, and uses a GNNExplainer-style explainer to output edge masks that are aggregated over k-hop neighbourhoods into node importance, thereby localising contiguous outcome-associated spatial regions and key proteins; it predicted progression in a 416-patient lung adenocarcinoma cohort (accuracy 0.68, F1 0.68, AUC 0.62, versus Ali et al. 0.61/0.59/0.57 and SPACE-GM 0.55/0.52/0.