This repository provides publication-oriented analysis code for spatial studies of human lymph-node remodelling across age groups. It contains code only; no raw imaging data, H5AD objects, cell-level tables, source data, or figure exports are distributed.
- src/fig1: Fig. 1 scripts, including area mixing correction, Ki-67 analysis, and fixed-15-um/fixed-128-cell T-cell dispersion analysis.
- src/fig2: Fig. 2 scripts for paracortical distance and T-cell subtype analyses.
- src/fig3: Fig. 3 scripts for neighbourhood, distance, and exclusion analyses.
- src/extended_data: Extended Data Fig. 2 scripts.
- environment.yml: reproducible Conda environment specification.
- config.example.json: illustrative, non-sensitive path layout; scripts use command-line arguments rather than this file.
Create the analysis environment from the repository root:
conda env create -f environment.yml
conda activate 3dln-analysis
Use only appropriately approved and de-identified data. Scripts expect project-relative input directories such as data/deidentified/ and write outputs under an ignored results/ directory. Required columns, annotations, and H5AD structure are described in the figure-specific READMEs. Do not place protected or identifiable data in this repository.
Run commands from the repository root using module notation:
python -m src.fig1.fig1c_area_mixing_correction --shape-path <SHAPE_METRICS_CSV> --mixing-path <ZONE_MIXING_CSV> --output-dir <OUTPUT_DIR>
python -m src.fig1.run_fig1_public_pipeline --project-dir . --output-dir results/fig1
python -m src.fig1.fig1h_calculate_dispersion_fixed15_128 --data-dir <DATA_DIR> --out-dir <RAW_OUTPUT_DIR>
python -m src.fig1.fig1h_adjust_summarize_and_test --calc-dir <RAW_OUTPUT_DIR> --out-dir <RESULTS_DIR>
python -m src.fig2.run_fig2_public_pipeline --project-root .
python -m src.fig3.run_fig3_public_pipeline --project-root .
python -m src.extended_data.ext2a_disorder_scores --project-dir .
python -m src.extended_data.ext2b_nn_distance --project-dir .
Each figure directory contains a README with its expected inputs and outputs. The full analyses require controlled input data and are intentionally not run as part of this public release.
The repository makes analysis logic and fixed calculation parameters available. Raw multiplexed imaging, participant-linked metadata, cell-level matrices, and derived source-data tables remain subject to ethical approval, consent, and data-sharing restrictions. Reproduction requires access to approved, de-identified inputs with the documented schema; this repository alone cannot recreate protected data or manuscript figures.
No open-source license has been added. Without a license, others may view the repository but do not receive broad permission to reuse or redistribute it.