SenSASP: A Unified, Multi-Layer Database of Senescence and SASP Genes

A fully scripted, one-command-reproducible pipeline that unifies four senescence / SASP gene resources onto a single canonical identifier and enriches every gene with cross-species conservation, tissue/cell-type expression, and protein–protein interaction annotation.

Quick start

./main.sh            # install deps, run the data pipeline, regenerate all figures
./main.sh --no-figs  # data pipeline only (skip the notebook)

Everything is fetched from public APIs at run time — no manual downloads, no credentials. Total runtime is a few minutes (the Human Protein Atlas step makes one request per gene and dominates the wall time).

What it does

Figures are generated separately from the pipeline outputs by notebooks/figures.ipynb.

Running individual steps

PYTHONPATH=. python run_pipeline.py          # all steps
PYTHONPATH=. python run_pipeline.py 4 5 6     # only these (reuse earlier outputs)

Layout

senescence_db_pipeline/
├── main.sh                 # one-click reproduction
├── run_pipeline.py         # step runner
├── requirements.txt
├── pipeline/               # step1..step8 modules + config.py
├── notebooks/figures.ipynb # all figure-generation code
├── data/raw, data/processed
├── results/                # database + metrics + report
├── figures/                # generated PNGs
└── paper/paper.md          # manuscript (Intro/Methods/Data/Results/Discussion)

Data sources

CellAge & GenAge (HAGR), SenMayo (MSigDB SAUL_SEN_MAYO / M45803, Saul et al. 2022), Reactome Cellular Senescence (R-HSA-2559583); annotation via MyGene.info, Ensembl BioMart, GTEx v8, Human Protein Atlas, and STRING v12.

Web database (GitHub Pages)

The interactive database is served from docs/ via GitHub Pages:

https://xuan13hao.github.io/sensasp

To enable: Settings → Pages → Source: Deploy from branch → Branch: main → Folder: /docs

To regenerate site data after re-running the pipeline:

python3 generate_site_data.py

This rebuilds docs/data/genes.json, genes_full.json, metrics.json, and sources.json.

Key results

1,460 summed source entries → 1,250 unique genes (210 redundant collapsed, 14.4%); 173 genes confirmed by ≥2 sources, 2 (IL6, JUN) by all four. Net-new annotation reaches 95.8% (conservation), 97.8% (expression), 92.9% (interactions), with 89.3% complete across all three layers.