<<<<<<< HEAD

Scleroderma Prediction Hackathon Project

This repository contains code and resources for predicting scleroderma from patient data using machine learning and NLP, as well as a Shiny frontend for interactive predictions. For more information, below are the slides from the presentation at the hackathon:

📊 Presentation: "ScleroDx - A Diagnostic Algorithm WebTool"

📄 View Full Presentation (PDF)

Project Structure

• scleroderma_api.py: FastAPI backend for predictions and recommendations
• app.R: R Shiny frontend for user interaction
• requirements.txt: Python dependencies
• .gitignore: Excludes large data/model files from Git
• *.joblib: Model and preprocessing objects (not tracked by Git)
• *.csv, *.txt: Patient data files (not tracked by Git)

How to Use

1. Backend: Install Python dependencies (pip install -r requirements.txt), start FastAPI with python scleroderma_api.py.
2. Frontend: Open R, install required packages (shiny, httr, jsonlite), and run shiny::runApp('app.R').
3. Deployment: Deploy the Shiny app to shinyapps.io for a public frontend; deploy the backend to a public server if needed.

Notes

• Sensitive patient data and large model files are excluded from version control.
• Update API URLs in app.R if deploying backend elsewhere.

---

For more details or questions, contact the repo owner.

Hackathon

4d57d00c2c5a385151a5bf25c3776e9c0e16e7ee