https://github.com/tomasonjo/clinical-agent

This project is designed for a simple LLM agent with tools on a clinical knowledge graph in Neo4j.
https://github.com/tomasonjo/clinical-agent

agent chatgpt graph graphrag langchain llm neo4j openai

Last synced: about 1 month ago
JSON representation

This project is designed for a simple LLM agent with tools on a clinical knowledge graph in Neo4j.

• Host: GitHub
• URL: https://github.com/tomasonjo/clinical-agent
• Owner: tomasonjo
• License: apache-2.0
• Created: 2024-06-06T17:55:42.000Z (7 months ago)
• Default Branch: main
• Last Pushed: 2024-08-17T18:03:44.000Z (4 months ago)
• Last Synced: 2024-10-29T23:43:43.743Z (about 2 months ago)
• Topics: agent, chatgpt, graph, graphrag, langchain, llm, neo4j, openai
• Language: Python
• Homepage:
• Size: 443 KB
• Stars: 17
• Watchers: 2
• Forks: 6
• Open Issues: 1
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
# clinical-agent

This project is designed for a simple LLM agent with tools on a clinical knowledge graph in Neo4j.

![Workflow diagram](https://raw.githubusercontent.com/tomasonjo/clinical-agent/main/static/clinical.png)

## Setup

1. Download and restore the [Clinical database dump](https://drive.google.com/file/d/1r5mHYuDjl-ilB8_vqgCqMUnziqGbOete/view?usp=drive_link).

2. Set environment variables in `.env`.

3. Run the following command:

```

docker compose up 

```

## Example questions

*  What are associated foods with cancer?

*  What are known clinically relevant gene variants for gene CLN3?

*  In which tissues are expressed proteins associated with prostate carcinoma?

## Todo

* Better entity mapping from input to database (Using a combination of keyword search and/or biomedical text embeddings)

* More tools

## Contributions

Contributions are welcomed