PPI Shiny App

An interactive Protein-Protein Interaction (PPI) visualization tool built with Shiny for Python. This application allows users to explore interaction networks from high-quality datasets like BioPlex and HuRI.

• Interactive Network Visualization: Uses vis-network for high-performance graph rendering.
• Multiple Datasets: Supports BioPlex 293, BioPlex HCT116, and HuRI.
• Gene-Centric Exploration: Search for specific genes and visualize their first-degree interaction partners.
• REST API Endpoint: Programmatic access to generate subnetwork HTML fragments.
• Dockerized Deployment: Ready for production with Docker and Uvicorn.

├── app.py                # Main Shiny application & Starlette API
├── config.py             # Global configurations and data initialization
├── data_loader.py        # Logic for reading PPI datasets and neighbor retrieval
├── graph_utils.py        # NetworkX and Pyvis graph generation utilities
├── ui_modules.py         # Modular UI component definitions
├── server_modules.py     # Server-side reactive logic
├── update_data.py        # Script for data maintenance
├── generate_static_graph.py # CLI tool for static graph generation
├── Dockerfile            # Container configuration
├── deploy.sh             # Deployment automation script
├── data/                 # PPI dataset storage (TSV/CSV)
├── lib/                  # Local JS/CSS dependencies (vis-network, tom-select)
└── www/                  # Static assets and generated graph cache

As per the project guidelines, use conda to manage the environment and packages:

# Create and activate the environment
conda create -n bio_env python=3.12
conda activate bio_env

# Install dependencies from conda-forge
conda install -c conda-forge shiny pandas networkx pyvis uvicorn

Run the application locally using shiny:

shiny run app.py --reload

Or via uvicorn (as configured in the Dockerfile):

uvicorn app:app --host 0.0.0.0 --port 5070

The project includes a Dockerfile and deploy.sh for easy deployment.

chmod +x deploy.sh
./deploy.sh

The application will be accessible at http://localhost:5070/ppi.

The application exposes a subnetwork generation endpoint:

GET /graph/{dataset}/{gene}

Example: http://localhost:5070/ppi/graph/huri/TP53

This project is licensed under the MIT License.