This application is available on Nprod.net
The dataset is from the LOTUS initiative and Wikidata.
- Dash for its web ui.
- EPAM's Ketcher for the molecule editor.
- Rdkit for the molecule massaging.
docker-compose run -it backend python update.py # You do not have to run it everytime, but that's recommended until we have a schema versioning
docker-compose up --buildThe web server is then available on http://localhost:3000 and the API on http://localhost:5000.
make tests
To run it yourself, the source is available at: https://github.com/lotusnprod/lotus-search:
- Install dependencies using poetry
- If you do not have poetry installed:
curl -sSL https://install.python-poetry.org | python3 -
- Then:
poetry self update && poetry update && poetry install
- If you do not have poetry installed:
- Run
python update.py(takes a few minutes) - Run
uvicorn main:app --reload(almost instant)
Adriano Rutz, is the one that pushed me into that, we are both part of the team behind LOTUS.
- Personal website: https://www.bjonnh.net
- You can find me on mastodon too(t): https://mastodon.social/@bjonnh
- Personal website: https://adafede.github.io
- And on mastodon: https://mastodon.online/@adafede
Your molecules are never stored unless they make our service crash.
But if your molecules are super secret, it is like your extremities, don't insert them in machines you don't know or understand.
Also, this is an experimental tool meant to test things, you're not supposed to rely on it for anything important, and it comes with no warranty or support whatsoever.
https://raw.githubusercontent.com/lotusnprod/lotus-search/main/LICENSE