1025-Path2Space: AI-Predicted Spatial Transcriptomics

Paper Talk · May 24, 2026 · 24m

The research introduces Path2Space, a deep-learning model designed to predict spatial gene expression directly from standard H&E histopathology slides. By eliminating the need for expensive molecular assays, this tool allows researchers to analyze the tumor microenvironment at scale using routine clinical images. The authors demonstrate that the model accurately identifies three prognostic subgroups, termed SpatioTypes, which correlate with distinct survival outcomes in breast cancer patients. Furthermore, the model discovers spatial biomarkers that predict patient response to chemotherapy…