Learning Phenotype Associated Signature in Spatial Transcriptomics with PASSAGE (Small Methods 5/2025)

Spatial Signature Identification The cover image depicts a novel deep learning framework, PASSAGE, designed to automatically identify phenotype-associated spatial signatures across large-scale heterogeneous spatial transcriptomics cohorts, schematically represented by intertwining passages connecting different puzzle pieces (spatial transcriptomics record the ground-truth spatial location of cells, a process akin to reconstructing a scattered puzzle into its original landscape). More in article number 2401451 by Guangdun Peng, Zhi-Jie Cao, and Ge Gao and co-workers.