Medical Drug Interaction and Analysis API

This project provides a Medical Drug Interaction and Analysis API, allowing users to extract detailed medication-related information from multiple online sources, analyze drug interactions, and manage this data via a backend API.

---

Table of Contents

• Overview
• Key Features
• Project Structure
• Setup Instructions
• Endpoints
• Technical Details
• Contributing

---

Overview

The API integrates:

1. Web scraping from trusted medical resources (e.g., drugs.com, 1mg.com).
2. Natural language processing for content extraction and categorization using ChatGroq and LangChain.
3. Persistent storage and querying of drug metadata using ChromaDB.
4. RESTful endpoints for user interaction with medication data.

This project is built using Python with the Flask web framework and supports asynchronous background task execution via threading.

---

Key Features

1. Drug Data Extraction

   • Scrapes drug information from reliable sources.
   • Uses a sophisticated language model (ChatGroq) for text processing and data extraction.

2. Drug Interaction Analysis

   • Identifies potential interactions between drugs.
   • Records metadata and interaction details for easy querying.

3. User Medication Management

   • Links user-provided medication data with detailed drug analysis.

4. Persistent Storage

   • Stores drug metadata and interaction data in ChromaDB.
   • Enables fast querying of existing records.

5. REST API

   • Provides endpoints to trigger background tasks and fetch/update medication data.

---

Project Structure

project/
├── Files/
│   ├── Original.py           # Main Python File
│ # Background task handling
├── requirements.txt          # Python dependencies
└── README.md                 # Documentation

---

Setup Instructions

Prerequisites

• Python 3.8 or higher
• API key for ChatGroq
• Libraries: Flask, BeautifulSoup, requests, ChromaDB, LangChain

Installation

1. Clone the repository:

   git clone https://github.com/krishnaAg16/DailyDoseLLM.git
   cd DailyDoseLLM.git

2. Install dependencies:

   pip install -r requirements.txt

3. Configure ChatGroq API key:

   • Replace YOUR_API_KEY in app.py with your actual API key.

4. Run the Flask application:

   python app.py

---

Endpoints

1. Add Drug Data for a User

GET /addDrug/<usr>

• Description: Triggers a background task to scrape, analyze, and store drug data for a specific user.
• Parameters:
  • usr: The user identifier.
• Response:
  • 202 Accepted: Task started successfully.
  • 500 Internal Server Error: Task initialization failed.

Example:

curl -X GET http://127.0.0.1:5000/addDrug/user123

---

Technical Details

1. Web Scraping

• Data Sources:
  • Drugs.com
  • 1mg.com
• Tools Used:
  • BeautifulSoup for HTML parsing.
  • requests for making HTTP requests.

2. Natural Language Processing

• LLM: ChatGroq (Llama-3.1-70B)
• Processing Steps:
  • Scraped text is passed to the ChatGroq model via the LangChain pipeline.
  • Extracted details include:
    • Drug Name
    • Recommended and avoided foods
    • Interacting drugs
    • Common and serious side effects
    • Treatment use cases

3. Database (ChromaDB)

• Purpose:
  • Store and query drug metadata.
• Data Model:

  {
    "name": "DrugName",
    "super_food": ["list of recommended foods"],
    "bad_food": ["list of foods to avoid"],
    "interactive_drug": [
      {"name": "InteractingDrug", "delta": "time in hours"}
    ],
    "common_symptom": ["list of common side effects"],
    "serious_symptom": ["list of serious side effects"],
    "treatment": ["medical conditions treated"]
  }

4. Asynchronous Tasks

• Background tasks are handled using Python's threading module.
• Tasks include:
  • Fetching and analyzing user medication data.
  • Updating drug interactions in the database.

---

Contributing

Contributions are welcome! If you'd like to contribute:

1. Fork the repository.
2. Create a new branch for your feature or bug fix.
3. Submit a pull request with a detailed description of changes.

---