Appropriateness of scRNA-seq-inspired normalization for SRT

Determine whether scRNA-seq-derived normalization practices, particularly library-size normalization, are uniformly appropriate across diverse spatially resolved transcriptomics platforms and data types where observations may capture inconsistent biological units.

Background

Normalization procedures from scRNA-seq pipelines, especially library-size normalization, are widely applied to SRT data. However, SRT observations can encapsulate varying biological content (e.g., multiple cells, extracellular space), so library size can reflect biological rather than purely technical variation.

Applying such normalization indiscriminately risks overcorrection and loss of biological signal, affecting tasks like spatial clustering, differential expression, and detection of spatially variable genes. Clarifying when these practices are valid across SRT modalities is critical for robust integrative analyses.