Awesome

DivGAN

DivGAN uses generative adversarial networks to perform small molecule map generation tasks, which are implemented in the Keras framework. It allows the user to run the model to generate a reference set of drug-like molecules.

Requirement

Refer to requirement.txt

Installation

• Install python 3.7 in Linux and Windows.
• If you want to run on a GPU, you will need to install CUDA and cuDNN, please refer to their websites for corresponding versions.
• Add your installation path and run the following command to install the DivGAN libraries in one step

pip install -r requirement.txt

Running DivGAN

You need to open main.py, run load_weights to read the pre-trained weights and get the generated molecules. Or provide training set molecules into graph coding for model training.