mage of a breast cancer tissue sample overlaid with visual representations of spatially interacting patterns of tumor and immune activity obtained through latent space factorization of 10X Visium spatial transcriptomics. The first overlay represents the spatially varying influence of the tumor (red) and immune (blue) patterns and highlights the hotspots of tumor and immune activity as well as of their interaction. The scatterpie representation in the second overlay represents multiple interacting biological processes in a single visualization. The SpaceMarkers algorithm identifies regions of interaction between the tumor and immune patterns and detects molecular changes associated with tumor-immune interaction.