Data Preparation for SpatialCOC
After configuring the SpatialCOC environment, please download the datasets from the Zenodo repository and extract it to the “Data” subfolder.
To comprehensively evaluate the integration capability of SpatialCOC, we systematically examined the inherent spatial structural heterogeneity in real tissues, including:
Smooth gradients: e.g., continuous gene expression changes across the cerebral cortex.
Sharp boundaries: e.g., abrupt transitions between the hippocampus and cortex.
Hierarchical structures: e.g., zonal organization in pancreatic germinal centers or thymic lobules.
Complex regions: e.g., developing liver and heart in mouse embryos.
Additionally, we incorporated authentic technical noise interference to ensure that our evaluation of SpatialCOC’s integration performance reflects the complexity of real tissue environments, including:
Measurement errors introduced by limitations in spatial resolution.
Gene dropout events caused by fluctuations in molecular capture efficiency.
Unavoidable batch effects across different experimental batches or sample preparation procedures.
