Longevity Simulator
A browser-based tool showing how curing diseases or slowing aging could change lifespan and healthspan for a U.S. resident. It's a static site with no backend: everything runs in the browser.
Data sources:
- CDC/NCHS United States Life Tables (2010–2019): mortality rates by age
- CDC Multiple Cause of Death database (2024): cause-of-death fractions by age
- IHME Global Burden of Disease (2023): chronic-condition prevalence by age
Run locally
cd web && python3 -m http.server 5180
# open http://127.0.0.1:5180
Any static file server works. Opening index.html directly from disk won't, because the page fetches data.json.
Deploy
Deploy web/ as a static site (on Vercel, set the project's Root Directory to web, with no framework and no build command).
Project structure
web/ # The app (deployed as-is)
index.html # Layout and copy
app.js # UI wiring, charts (Plotly.js), CSV export
model.js # The simulation: cause removal, aging slowdown, survival, healthspan, Gompertz fit
model.check.mjs # Regression check for model.js: node web/model.check.mjs
data.json # Bundled runtime data, generated from data/
data/ # Processed CSVs (committed)
preprocessing/ # Python scripts: raw_data/ -> data/ -> web/data.json
raw_data/ # Large raw source files (gitignored)
Updating the data
pip install -r requirements.txt
cd preprocessing && python process_life_table.py && python process_mmcd.py && cd ..
python preprocessing/export_web_data.py # rebuilds web/data.json
node web/model.check.mjs # reference values may need updating if the data changed
See preprocessing/README.md for details on the raw data.
License
MIT License, see LICENSE.
Author
Alexander Epstein, Cao Lab, Rockefeller University
