STRAND: a comprehensive subcellular-resolved spatial transcriptome RNA architecture and navigation database

Subcellular RNA localization serves as a critical regulatory layer governing cellular polarization, local translation, and functional compartmentalization. Recent advances in high-resolution spatial transcriptomics have provided unprecedented opportunities to resolve these intracellular architectures at single-molecule resolution. However, relevant datasets are dispersed across studies and repositories, vary substantially in processing procedures and standards, severely limiting their cross-study reuse. To bridge this gap, we present STRAND (Subcellular-Resolved Spatial Transcriptome RNA Architecture and Navigation Database), a curated data resource and analysis platform for the automated harvesting, quality control, and interpretation of subcellular spatial transcriptomics data. Through an agent-assisted curation pipeline, STRAND has assembled 35 subcellular-resolution datasets from six major platforms, covering 17 tissue types in human and mouse and totalling ∼4.9 billion transcript coordinates. Building upon this standardized data resource, STRAND provides an analytical workflow spanning cell segmentation evaluation, cell-type annotation, RNA localization/colocalization pattern annotation, and subcellular compartment inference. Two case studies demonstrate that STRAND successfully identifies established RNA localization patterns in cultured cell lines and detects drug-induced RNA relocalization in cardiomyocytes. STRAND is openly available at https://strand.phasep.pro.