Update README.md

README.md

@@ -10,14 +10,121 @@ pinned: false
 short_description: Research
 ---
 
-# 🧬 Ultimate AI-Powered Research MCP (Hugging Face Space)
+MedGenesis AI 🧬
+Unified Biomedical Research Assistant—AI, Literature, Safety & Knowledge Graphs
 
-- 🔬 Unified search: arXiv, PubMed, UMLS, OpenFDA, OpenAI
-- 🧠 AI synthesis & Q&A: GPT-4o powered
-- 🩺 Biomedical enrichment: Drugs, CUIs, adverse events
-- 🚀 Plug-and-play: Add more APIs in minutes!
+🚀 Overview
+MedGenesis AI is the next-generation biomedical research assistant that unifies PubMed, arXiv, UMLS, OpenFDA, and advanced AI (OpenAI GPT-4o) into one seamless platform.
+Discover, synthesize, and visualize research, drugs, concepts, and safety—all in a single app.
+
+Semantic search across top biomedical databases
+
+UMLS concept mapping and standardization
+
+Drug safety insights powered by OpenFDA
+
+AI-powered summaries and follow-up Q&A
+
+Interactive knowledge graph for instant relationship exploration
+
+Export results as PDF or CSV
+
+Workspace to save and revisit your research sessions
+
+✨ Key Features
+Unified Search: Find and aggregate results from PubMed and arXiv, ranked by semantic similarity.
+
+Biomedical Concept Augmentation: Map results to UMLS concepts for standardized knowledge.
+
+Drug & Safety Insights: Connect literature with latest FDA safety data.
+
+AI-Powered Synthesis: Summarize, answer questions, and suggest next research steps using GPT-4o.
+
+Knowledge Graph: Visual, interactive network of papers, drugs, and concepts.
+
+Export & Workspace: Download data, save sessions, and organize your findings.
+
+🖥️ Live Demo
+Try the app on Hugging Face Spaces
+
+🐳 Docker-Based Deployment
+MedGenesis AI uses a Dockerfile for guaranteed reproducibility and robust dependency management.
+
+Run Locally:
+bash
+Copy code
+git clone https://huggingface.co/spaces/YOUR_SPACE_URL
+cd YOUR_REPO
+docker build -t medgenesis-ai .
+docker run -p 7860:7860 medgenesis-ai
+Then open http://localhost:7860 in your browser.
+
+On Hugging Face Spaces:
+Make sure your Dockerfile, requirements.txt, and app.py are present in your repo root.
+
+Push all code to your Space.
+
+In the Space settings, select "Docker" as the runtime.
+
+Your Space will build and launch automatically.
+
+⚡ Quick Start (Streamlit UI)
+Enter a biomedical research question.
+
+Click Run Search to query all sources.
+
+Explore:
+
+Results: Detailed, AI-synthesized findings.
+
+Knowledge Graph: Interactive visualization.
+
+Visualizations: Year histogram and more.
+
+Download results or save to your workspace.
+
+Ask follow-up questions with the integrated AI assistant.
+
+🛠️ Customization
+Add new data sources by extending mcp/ utilities.
+
+Update color/theme in .streamlit/config.toml.
+
+Want to run in SDK mode? Simply remove the Dockerfile and deploy with requirements.txt as usual.
+
+Use the built-in workspace to organize your research projects.
+
+🤝 Contributing
+Pull requests, bug reports, and feature suggestions are welcome!
+For large features, open an issue first to discuss your ideas.
+
+🧑‍💻 Author
+Oluwafemi Idiakhoa
+
+📝 License
+MIT License
+
+❤️ Acknowledgments
+Hugging Face Spaces
+
+PubMed, arXiv, OpenFDA, UMLS
+
+OpenAI GPT-4o
+
+The open-source biomedical AI community
+
+🔗 Links
+MedGenesis AI Space
+
+OpenFDA
+
+UMLS
+
+arXiv
+
+PubMed
+
+Note on Docker
+If you’re using Docker, all dependencies (including Jinja2, pyvis) are installed as specified in the Dockerfile and requirements.txt—no extra setup needed.
+If deploying with Hugging Face SDK mode, ensure your requirements.txt matches your dependencies.
 
-**Run locally:**
-```bash
-pip install -r requirements.txt
-streamlit run app.py