Skip to content

smb-h/de-novo-drug-discovery

Repository files navigation

De Novo Drug Design

This project explores innovative approaches in drug discovery using recurrent and deep neural networks. The goal is to generate novel drug formulas with specialized structures, demonstrating the potential of artificial intelligence to expedite the identification of promising compounds for pharmaceutical development.

Getting Started

Installation

pip install -r requirements.txt

Preprocessing

python -m drug_design.data.preprocessing_molinf

Training the Model

python -m drug_design.models.train

Checking Accuracy & Loss

tensorboard --logdir reports/<date>/<experiment_name>/logs

Prediction

python -m drug_design.models.predict

Fine-tuning the Model

python -m drug_design.models.fine_tune

Contribution

  • SMBH

License

This project is licensed under the MIT. See the LICENSE file for details.