Spatial cell-cell communication inference based on cell-spot-ligand-receptor heterogeneous graphs.

SpaHCC uncovers spatially coherent cell-cell communication signals with significant biological insights in disease tissues, including Alzheimer’s disease, by integrating multi-modal data.

• Why it matters: Understanding cell communication in tissue microenvironments is crucial for insights into disease mechanisms and therapeutic targets, yet current methods struggle with spatial resolution and biological interpretability.
• What they did: SpaHCC employs a heterogeneous graph learning approach that combines single-cell and spatial transcriptomics with ligand-receptor knowledge, analyzing molecular features, spatial context, and cell states across multiple datasets.
• The result: The framework successfully identifies recurrent, biologically meaningful communication patterns, such as synaptic remodeling modules in Alzheimer’s disease, enabling improved interpretation of spatial cell interactions in complex diseases.