A spatially resolved transcriptomic and cellular landscape of colitis-associated colon cancer

Spatial transcriptomics datasets (Xenium Prime and Xenium v1) generated from colitis-associated colon cancer patient samples. Processed data are provided as .h5ad files.CAC_XeniumPrime_integrated.h5ad - Xenium Prime 5K data (default segmentation) with Novae computed niches and Novae spatial niches UMAP CAC_XeniumPrime_integrated_HarmonyUMAP.h5ad - Xenium Prime 5K data (default segmentation) with Novae computed niches with Harmony UMAP CAC_Xenium477_Harmony.h5ad - Xenium v1 dataset with default segmentation. CAC_Xenium477_Proseg_Harmony.h5ad - Xenium v1 dataset resegmented with Proseg v2.0.5 AnnData object summary: - Observations (`.obs`): • group: patient IDs • Subgroup: patient subsets based on SPP1 status • condition: disease type (CAC, UC, HC) • Broad cell types: `level1_abbvr` → 'Epithelial', 'Lymphoid', 'Myeloid', 'Stromal' • More detailed cell types: `level2rev_abbvr` → e.g., Epithelial, T, B, Macro, DC, Glial • Fine-grained cell subtypes: `level3rev_abbvr` Computed features: • Novae cellular neighborhoods: `novae_domains_7_corr`