# HGCL-DR

**Repository Path**: cool_c0de/HGCL-DR

## Basic Information

- **Project Name**: HGCL-DR
- **Description**: No description available
- **Primary Language**: Unknown
- **License**: Not specified
- **Default Branch**: main
- **Homepage**: None
- **GVP Project**: No

## Statistics

- **Stars**: 0
- **Forks**: 0
- **Created**: 2025-10-31
- **Last Updated**: 2025-10-31

## Categories & Tags

**Categories**: Uncategorized

**Tags**: None

## README

# HGCL-DR: A Heterogeneous Graph Contrastive Learning Framework for Drug Repositioning

Drug repositioning aims to identify new therapeutic uses for existing drugs, providing a cost-effective alternative to traditional drug development. This repository contains the official implementation of **HGCL-DR**, a novel **heterogeneous graph contrastive learning** framework that effectively integrates both **global** and **local** feature representations for drug repositioning.

---

## 🌟 Highlights

- **Improved Heterogeneous Graph Contrastive Learning**: Captures semantic consistency across heterogeneous relations.
- **Subgraph-Based Local Feature Learning**: Uses a bidirectional GCN and graph diffusion to model long-range dependencies.
- **Global-Local Fusion**: Contrastive learning enhances representation quality across different spaces.
- **Strong Performance**: Outperforms state-of-the-art baselines on four benchmark datasets in AUPR, AUROC, and F1-score.

---

## 🧠 Model Architecture

<!-- Insert your model diagram here -->
<!-- For example: ![HGCL-DR Architecture](./figures/hgcl_dr_model.png) -->
<p align="center"><img src="model (2).jpg" width="700"></p>

---

## Requirements

The latest tested environment for running this project includes the following:

- Python 3.8.1  
- PyTorch 1.8.1  
- DGL 0.8.1

## ▶️ How to Use

To train the HGCL-DR model on your dataset, simply run the following command:

```bash
python main.py