gentel-lab/origenemcp

开发工具

本页由 AI 翻译，技术细节请以原文为准。查看原文

40 stars 5 forks 质量 51 趋势 12

Repository Install

OrigeneMCP是OriGene生态系统的核心组成部分，OriGene是一个自我演化的多智能体系统，旨在扮演虚拟疾病生物学家的角色。

概览

OrigeneMCP是OriGene生态系统的核心组件，OriGene是一个自我演进的多智能体系统，旨在扮演虚拟疾病生物学家的角色。OriGene为生物医学人工智能研究提供统一平台，并于2025年WAIC大会上正式发布。纸质 |OriGene 代码 |OrigeneMCP 代码 | 最新更新：- 🚀 更新新工具：DepMap等- 🚀 OriGene于WAIC 2025正式发布！- 🔥 将OrigeneMCP集成到Bohrium AI for Science平台。- 🔥 将OrigeneMCP集成到CAMEL中。- 🚀 OrigeneMCP正式开源！

概述（AI 翻译）

OrigeneMCP是OriGene生态系统的核心组成部分，OriGene是一个自我演化的多智能体系统，设计为虚拟疾病生物学家。OriGene提供了一个统一的生物医学人工智能研究平台，并于2025年WAIC大会上正式发布。

|产品 | 论文 | 奥里吉恩代码 | OrigeneMCP 代码 |

1.OrigeneMCP 概述

最新动态：

2026-02-24 🚀 更新新工具：DepMap等。

2025-07-28 🚀 OriGene 在 WAIC 2025 正式发布！

2025-07-22 🔥 将OrigeneMCP集成到Bohrium AI for Science平台。

2025-07-19 🔥 将OrigeneMCP集成到CAMEL中。

2025-07-18 🚀 OrigeneMCP 正式开源！

本页由 AI 翻译，技术细节请以原文为准。

README

以下为原文 README 节选，完整内容请查看原文。

OrigeneMCP

OrigeneMCP is a core component of the OriGene ecosystem, a self-evolving multi-agent system designed to act as a virtual disease biologist. OriGene provides a unified platform for biomedical AI research and was officially launched at the 2025 WAIC.

1. OrigeneMCP Overview

Recent updates:

2026-02-24 🚀 Update new tools: DepMap, etc.

2025-07-28 🚀 OriGene officially launched at WAIC 2025!

2025-07-22 🔥 Integrated OrigeneMCP into the Bohrium AI for Science platform.

2025-07-19 🔥 Integrated OrigeneMCP into CAMEL.

2025-07-18 🚀 OrigeneMCP officially open-sourced!

OrigeneMCP is the MCP toolkit for the OriGene project that integrates over 600 tools and multiple databases (including ChEMBL, PubChem, FDA, OpenTargets, NCBI, UniProt, PDB, Ensembl, UCSC, KEGG, STRING, TCGA, Monarch, ClinicalTrials, and more) into an advanced integrated MCP server platform, combining disease biology and drug discovery tools to enable comprehensive multi-dimensional information retrieval across small molecules, proteins, genes, diseases, and other biological entities, serving as a unified interface for accessing and analyzing complex biomedical data to accelerate therapeutic research and development.

2. Deploy OrigeneMCP

Clone the repository:

git clone https://github.com/GENTEL-lab/OrigeneMCP.git

The dependencies of this repository are managed by uv. Before deployment, please make sure that uv is already installed. Use the following commands to install uv:

pip install uv

uv venv .venv --python=3.13

source .venv/bin/activate

Initialize the dependencies of this repository:

uv sync

Deploy the mcp service:

cp default.conf.toml local.conf.toml
# get your tavily_api_key in https://tavily.com/
# get your jina_api_key in https://jina.ai/

export PYTHONPATH=`fab pypath`
uv run -m deploy.web

NOTE:

Please ensure that port 8788 is not occupied. If it is occupied, please modify the port in file local.conf.toml to another port;
If you need to use Tavily search or Jina search, please configure the tavily_api_key or jina_api_key in file local.conf.toml;

3. Use OrigeneMCP

3.1. Use locally deployed OrigeneMCP

First, connect to OrigeneMCP and get all available tools.

from langchain_mcp_adapters.client import MultiServerMCPClient

# Choose the packages you want to use
tool_packages = [
    "chembl", "kegg", "string", "search", "pubchem", "ncbi",

在 GitHub 查看完整 README

安装

This server does not publish a one-line install command.

Open the repository installation guide 新手？先看 mcpskill.org 教程 →

配置

import asyncio

from camel.agents import ChatAgent
from camel.toolkits import OrigeneToolkit

# Choose the packages you want to use
tool_packages = [
    "chembl", "kegg", "string", "search", "pubchem", "ncbi",
    "uniprot", "tcga", "ensembl", "ucsc", "fda_drug", "pdb",
    "monarch", "clinicaltrials", "dbsearch", "opentargets"
]
tool_packages = [server+'_mcp' for server in tool_packages]

config = {
    "mcpServers": {
        package: {
            "transport": "streamable_http",
            "url": f"http://127.0.0.1:8788/{package}/mcp/"
        } for package in tool_packages
    }
}

async def main():
    # Use async context manager for automatic connection management
    async with OrigeneToolkit(config_dict=config) as origene_toolkit:
        user_msg = "what can you do?"
        agent = ChatAgent(
            "You are named origene assistant.",
            model="gpt-4o",
            tools=[*origene_toolkit.get_tools()],
        )
        response = agent.step(user_msg)
        print(response.msgs[0].content)

asyncio.run(main())