SpatialZoomer: multi-scale feature analysis of spatial transcriptomics.

SpatialZoomer extracts multi-scale spatial molecular features from spatial transcriptomics, automatically identifying critical scales and detecting disease-progression signals in Alzheimer’s at specific resolutions.

• Why it matters: Single-cell spatial transcriptomics needs methods that capture structure across multiple spatial ranges, since cellular interactions and organization vary by scale.
• What they did: SpatialZoomer uses a spectral graph approach with low-pass filters to derive features at multiple resolutions, then partitions a cross-scale similarity map to find critical scales.
• The result: It identifies Alzheimer’s disease progression signals at specific scales and reveals spatially heterogeneous niches and spatial context-dependent cell subtypes in tumour microenvironments, with comparable or superior performance and high computational efficiency.

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