Documentation.md

OntoGraph - Drug Interaction Analysis System

A sophisticated Python-based application that combines ontology reasoning with LLM capabilities to analyze drug interactions, conflicts, and alternatives. The system provides comprehensive analysis through both CLI and web interfaces.

Key Features

• Dual Interface Options:
  • Web-based interface using Streamlit
  • Command-line interface for direct interaction
• Comprehensive Drug Analysis:
  • Drug interaction detection
  • Conflict identification
  • Similar drug suggestions
  • Alternative medication recommendations
• Advanced Technology Stack:
  • RDF/OWL ontology for knowledge representation
  • Groq LLM integration for natural language processing
  • Agent-based architecture using Langgraph for modular analysis
  • Structured logging system

Technical Requirements

• Python 3.x
• Groq API key
• Required Python packages:
  • rdflib
  • langchain
  • python-dotenv
  • langchain-community
  • langchain-groq
  • langgraph
  • streamlit

Installation

1. Clone the repository:

git clone https://github.com/Manithj/onto_graph.git
cd onto_graph

2. Install dependencies:

pip install -r requirements.txt

3. Configure environment:
   • Create a .env file in the project root
   • Add your Groq API key:

GROQ_API_KEY=your_api_key_here

Usage

Web Interface

streamlit run app.py

Navigate to the displayed local URL to access the web interface.

Command Line Interface

python appcli.py

Project Structure

onto_graph/
├── app.py              # Streamlit web interface
├── appcli.py          # Command line interface
├── agents/            # Agent implementations
│   ├── __init__.py
│   ├── alternative_agent.py
│   ├── base_agent.py
│   ├── conflict_agent.py
│   ├── interaction_agent.py
│   └── similarity_agent.py
├── analyzers/         # Analysis coordination
│   ├── __init__.py
│   └── drug_interaction_analyzer.py
├── ontology/         # Knowledge base
│   └── DrugInteraction.owl
└── requirements.txt  # Project dependencies

Features in Detail

Agent System

• Base Agent: Common functionality for ontology queries
• Alternative Agent: Identifies alternative medications
• Similarity Agent: Finds similar drugs
• Conflict Agent: Detects drug conflicts
• Interaction Agent: Analyzes drug interactions

Analysis Pipeline

1. User input processing
2. Ontology querying
3. LLM-powered result synthesis
4. Structured response generation

Contributing

We welcome contributions! Please follow these steps:

1. Fork the repository
2. Create a feature branch
3. Commit your changes
4. Push to your branch
5. Create a Pull Request

Logging

The system maintains detailed logs in app.log, capturing:

• Information level messages
• Warning and error states
• Query execution details
• System state changes

License

This project is licensed under the Apache 2.0.