Repository containing code used to analyze single-cell RNA-seq data published in
Raw single-cell data available under GEO accession number GSE322796
Per-sample CellChat data (prior to modeling) is available on FigShare: LR Interaction Tables
- R was used for the core single-cell workflow (QC through CAR+ mapping and CellChat).
- Python was only used for downstream modeling of the per-sample CellChat outputs (per-interaction regression, rank aggregation, and EDF7 and EDF8 figures).
Single-Cell Analysis
 Pathway Enrichment 
 Visualisation 
|
Cell Frequency
 Cell-Cell Communication R  Python 
|
CART-Enterocolitis-Analysis-Scripts/
├── R Scripts/
│ ├── 01_QC.Rmd # Ambient RNA removal, doublet detection, QC filtering
│ ├── 02_Integration.Rmd # Batch correction
│ ├── 03_Clustering.Rmd # Global clustering, annotation, condition UMAPs
│ ├── 04_Subclustering_Stroma.Rmd # Sub-clustering workflow (stromal shown; same workflow applied to T cells, myeloid, B cells, plasma cells, epithelial, and endothelial subsets)
│ ├── 05_DEGs_and_GSEA_Stroma_Condition.Rmd # Condition-wise DEGs + GSEA
│ ├── 06_CLR_CellFreq_Stroma.Rmd # CLR cell-frequency analysis + forest plots (stromal shown; same workflow applied globally and to all other subsets)
│ ├── 07_Figs_Stroma.Rmd # Figures for stromal subset (bubble plots, stacked bars, condition plots, etc.)
│ └── 08_CAR_Mapping_and_Figs.Rmd # Certomics CAR+ cell mapping onto T cell subset, heatmaps, pie plots, etc.
│ └── 09_CellChat.Rmd # Sample-wise CellChat that was aggregated and analyzed further using regression model
│ └── sessionInfo.md # Full R session information and package versions
├── Python Scripts/
│ ├── colitis_analysis_script.ipynb # End-to-end CellChat downstream analysis notebook
│ └── clplots/ # Plotting helpers (volcano, circle, RRA/regression boxplots, etc.)
│ └── clutils/ # Data preparation, ranking, summarization, and regression utilities
│ └── package_versions.txt # Full Jupyter session information and package versions
└── README.md