Investigation of chemical structure recognition by encoder-decoder models in learning progress.
Please cite the following if you use any contents of this repository.
- DOI: 10.1186/s13321-023-00713-z
- Shumpei Nemoto
- main contributor
- Tadahaya Mizuno
- correspondence
git clone https://github.com/Nemoto-S/EDmodel.git
Please clone this repository locally and use it.
- descriptor generation
import sys
sys.path.append("EDmodel")
from src.trainer import Generator
from src.models import Config
gen = Generator()
gen.load(Config(),"models/model_94.pth")
gen.get_descriptor(input_SMILES) # input_SMILES: List of your smiles
- structure generation
gen.generate(descriptor) # descriptor: np.array or torch.tensor
If you have any questions or comments, please feel free to create an issue on github here, or email us:
- 88nemo77[at]gmail.com
- tadahaya[at]gmail.com
- lead contact