Drug Discovery with Graph Neural Networks

This folder contains the code and resources for GNN Drug Discovery project.

Table of Contents

• Introduction
• Installation
• Usage
• Contributing
• License
• Credits

Introduction

In this project, we aim to develop a Graph Neural Network (GNN) model for drug discovery. The goal is to leverage the power of GNNs to predict the efficacy and properties of potential drug candidates. See conference poster for details.

Installation

To run the code in this repository, you need to have the following dependencies installed:

• Python 3.7 or higher
• PyTorch
• RDKit
• ...

You can install the required dependencies by running the following command: pip install -r requirements.txt

Credits

Authors: Mihir S. Arya, Isabella Chittaro, Rinny Fan, George Simmons

We would like to thank Dr. Cristian Minoccheri, Emily Wittrup, and the University of Michigan Undergraduate Research Opportunity Program for their relentless support throughout this project.

Biomedical & Clinical Informatics Lab - University of Michigan, Ann Arbor