https://github.com/cx-luo/rxn-yield-prediction
A Deep Learning model for predicting chemical reaction yields, enabling faster reaction optimization and synthesis planning.
https://github.com/cx-luo/rxn-yield-prediction
aidd drug-discovery
Last synced: 8 months ago
JSON representation
A Deep Learning model for predicting chemical reaction yields, enabling faster reaction optimization and synthesis planning.
- Host: GitHub
- URL: https://github.com/cx-luo/rxn-yield-prediction
- Owner: cx-luo
- Created: 2025-09-26T08:35:01.000Z (9 months ago)
- Default Branch: main
- Last Pushed: 2025-09-26T15:35:02.000Z (9 months ago)
- Last Synced: 2025-09-26T17:37:32.022Z (9 months ago)
- Topics: aidd, drug-discovery
- Language: Python
- Homepage:
- Size: 817 KB
- Stars: 0
- Watchers: 0
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
### Reaction Yield Predictor Demo
This directory contains a simple reaction yield prediction demo using PyTorch and RXNFP fingerprints.
#### Conda Environment
```bash
conda env create -f environment.yml
conda activate helsinki-pre-assignment
```
#### Files
- `yield_predict.py`: Model and training loop
- `rxn_data_processing_utils.py`, `rxn_logger.py`: Data utilities and logging
- `test_organic_reactions.csv`: Example dataset
#### Run
```bash
python yield_predict.py
```
Notes
- The script downloads/loads RXNFP model weights on first run.
- Training artifacts are saved to `yield_predictor.pth` every 10 epochs.
- For custom data, adapt column names to those used in `yield_predict.py`.