smoothclust: Spatial smoothing and boundary-density summaries for spatial transcriptomics

smoothclust provides a spatial-smoothing clustering workflow and a boundary density metric that quantifies label transitions on a spatial graph from only coordinates and domain labels.

• Why it matters: Spatial domain maps are widely produced, but standardized, interpretable sample-level measures of tissue architecture and domain organization are less established.
• What they did: The framework computes raw boundary density as the proportion of neighboring spatial locations assigned to different labels, and also reports relative boundary density by dividing raw BD by its expectation under random relabeling.
• The result: Using smoothclust, researchers can turn spatial domain label maps into local and sample-level summaries for exploratory analysis and cross-sample or cross-condition comparisons, with evaluations across multiple spatial transcriptomics platforms and simulations.

AI summary of the abstract · may contain errors · not medical advice · read the paper.