Metadata-Version: 2.1
Name: molbloom
Version: 0.1.5
Summary: Purchaseable SMILES filter
Home-page: https://whitead.github.io/molbloom/
Author: Andrew White
Author-email: andrew.white@rochester.edu
License: MIT
Classifier: Programming Language :: Python :: 3
Classifier: License :: OSI Approved :: MIT License
Classifier: Operating System :: OS Independent
Classifier: Topic :: Scientific/Engineering :: Chemistry
Classifier: Topic :: Scientific/Engineering :: Artificial Intelligence
Classifier: Typing :: Typed
Description-Content-Type: text/markdown
License-File: LICENSE

# molbloom

Can I buy this molecule? Returns results in about 500 ns and consumes about 100MB of RAM (or 2 GB if using all ZINC20).

```sh
pip install molbloom
```

```py
from molbloom import buy
buy('CCCO')
# True
but('ONN1CCCC1')
# False
```

If `buy` returns `True` - it may be purchasable with a measured error rate of 0.0003. If it returns `False` - it is not purchasable.
The catalog information is from ZINC20. Add `canonicalize=True` if your SMILES are not canonicalized (requires installing rdkit).

If you want to look at the broader catalog of all molecules that are not in stock:
```py
buy('CCCO', instock=False)
```
the reference for that is all ZINC20 from October 2021. *On first execution of `instock=False` it will download 2.0 GB of data to a cache directory.*
