ARCADE disentangles cell abundance from continuous transcriptional state in spatial transcriptomics

ARCADE disentangles cell-type abundance from continuous within-type transcriptional programs in spatial transcriptomics, improving deconvolution accuracy and stabilizing state inference across simulated and semi-synthetic benchmarks.

• Why it matters: Spatial spot measurements entangle cell abundance with transcriptional variation, and unstable single-cell references can further blur both signals.
• What they did: ARCADE builds a robust reference via hyperparameter optimization and consensus filtering, then infers composition first and models continuous intratype programs around that compositional anchor.
• The result: In matched human dorsolateral prefrontal cortex datasets with spatial proteogenomic readouts, ARCADE distinguishes localized cell enrichment from astrocyte and microglial activation programs, improving interpretability.

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