🧬 LigBind AI

AI-Powered Drug Discovery Platform

Predict molecular properties, perform molecular docking, and accelerate computational drug discovery through Machine Learning and Artificial Intelligence.

---

🚀 Overview

LigBind AI is an AI-powered drug discovery platform that combines Machine Learning, Chemoinformatics, and Molecular Docking into a single web application.

The platform enables researchers, students, and developers to:

• Predict molecular properties
• Estimate protein-ligand binding affinity
• Perform virtual screening
• Explore computational drug discovery through an intuitive interface

The current implementation focuses on QSAR modeling and AutoDock Vina integration, while the long-term goal is to evolve LigBind AI into a research platform powered by Graph Neural Networks, Protein Language Models, and Generative AI.

---

📸 Screenshots

Dashboard

---

Docking Page

---

✨ Features

Current Features

• 🔬 LogP Prediction
• 🧪 pIC50 Prediction
• ⚛️ Molecular Docking using AutoDock Vina
• 🧬 Protein Selection
• 📊 Docking Score Visualization
• 👤 User Authentication
• 🎨 Modern React Interface
• ⚡ FastAPI Backend

---

Planned Features

• 🚧 Graph Neural Networks (GNN)
• 🚧 ADMET Prediction
• 🚧 Protein Language Models
• 🚧 Diffusion Models
• 🚧 Deep Docking
• 🚧 Molecule Generation
• 🚧 AI Research Assistant
• 🚧 Multi-Agent Drug Discovery Pipeline

---

🏗️ System Architecture

flowchart TD

A[User]

--> B[React Frontend]

B --> C[FastAPI Backend]

C --> D[LogP Model]

C --> E[pIC50 Model]

C --> F[Docking Engine]

D --> G[Prediction Results]

E --> G

F --> G

Loading

---

🧠 Current AI Pipeline

flowchart LR

A[SMILES]

--> B[RDKit]

--> C[Morgan Fingerprint]

--> D[Random Forest]

--> E[LogP / pIC50]

--> F[AutoDock Vina]

--> G[Binding Affinity]

Loading

---

🛠 Tech Stack

Frontend

• React
• Vite
• Bootstrap
• Axios

Backend

• FastAPI
• Python

Machine Learning

• RDKit
• Scikit-learn
• NumPy
• Pandas

Docking

• AutoDock Vina
• Meeko

Database

• MongoDB

---

📂 Project Structure

LigBind-AI
│
├── backend
│   ├── ml_engine
│   ├── auth.py
│   ├── main.py
│   └── ...
│
├── frontend
│   ├── client
│   └── server
│
├── docs
│   └── screenshots
│
├── README.md
├── README_SETUP.md
├── CONTRIBUTING.md
├── LICENSE
└── .gitignore

---

🚀 Quick Start

Clone the repository

git clone https://github.com/Anas-progk/LigBind-AI.git

For complete installation instructions, please refer to:

📄 README_SETUP.md

---

🔬 Research Vision

LigBind AI is intended to grow into a comprehensive AI-assisted drug discovery platform.

The planned research directions include:

• Graph Neural Networks for molecular property prediction
• Protein Language Models for protein representation
• Diffusion Models for molecule generation
• Deep Learning-based docking prediction
• AI-assisted molecular design
• Multi-Agent AI workflows for drug discovery

---

🗺️ Development Roadmap

Version 1

• ✅ QSAR Prediction
• ✅ Molecular Docking
• ✅ Authentication

---

Version 2

• 🚧 Graph Neural Networks
• 🚧 ADMET Prediction

---

Version 3

• 🚧 Diffusion-based Molecule Generation
• 🚧 Deep Docking

---

Version 4

• 🚧 Protein Language Models

---

Version 5

• 🚧 AI-Assisted Drug Discovery Platform

---

🤝 Contributing

Contributions are welcome!

If you'd like to contribute:

1. Fork the repository
2. Create a new feature branch
3. Commit your changes
4. Open a Pull Request

Please read CONTRIBUTING.md before contributing.

---

⚠ Large Files

The following files are intentionally excluded because they exceed GitHub's file size limits:

• Trained ML models (*.pkl)
• ZIP backups
• Generated docking outputs

Please place these files in their appropriate directories before running the project.

---

⭐ Support

If you find this project useful,

please consider giving it a ⭐ on GitHub.

Contributions, ideas, and feedback are always welcome.

---

👨‍💻 Author

Anas

Engineering Student • AI Developer • Full-Stack Developer

---

📄 License

This project is licensed under the MIT License.