scp-kano-chebi

name

scp-kano-chebi

description

Use when you need to connect to the SciGraph SCP server for KANO-ChEBI / ElementKG-CHEBI (knowledge graph integrating element properties and functional group hierarchies from Periodic Table/Wikipedia/ChEBI to support KANO molecular representation learning and functional prompting) and call its MCP tools (query_cypher, get_kg_statistics, get_entity_details, get_experiment_workflow), including streamableHttp configuration with SCP-HUB-API-KEY and Python 3.10+ usage examples.

SCP-KANO-ChEBI (SciGraph) MCP client

What this SCP is

KANO-ChEBI (ElementKG-CHEBI) is a knowledge graph built to enhance molecular representation learning. It integrates chemical element properties and functional group hierarchies from the Periodic Table, Wikipedia, and the ChEBI ontology.

The graph serves as structured prior knowledge for the KANO model, enabling element-guided graph augmentation and functional prompting for molecular property prediction.

Connection info

• MCP server URL:
  • https://scp.intern-ai.org.cn/api/v1/mcp/37/SciGraph
• Auth header:
  • SCP-HUB-API-KEY: {API-KEY}

Install

pip install mcp

Configure (MCP config JSON)

{
  "mcpServers": {
    "SciGraph": {
      "type": "streamableHttp",
      "description": "这是一款面向科学研究的统一知识查询服务，集成了化学、生物等多个学科领域的知识图谱数据，支持跨学科知识检索、实体关系查询、领域知识问答等操作",
      "url": "https://scp.intern-ai.org.cn/api/v1/mcp/37/SciGraph",
      "headers": {
        "SCP-HUB-API-KEY": "{API-KEY}"
      }
    }
  }
}

Tools

query_cypher

Execute a Cypher query and return JSON results.

Arguments:

• cypher (string, required)
• kg_name (string|null, optional, default null)
• limit (int, optional, default 100)

Example arguments (KANO-ChEBI):

{
  "cypher": "MATCH (e:Experiment:KANO-ChEBI) RETURN e.id as experiment_id",
  "kg_name": "KANO-ChEBI",
  "limit": 5
}

get_kg_statistics

Return graph statistics.

Example arguments:

{ "kg_name": "KANO-ChEBI" }

get_entity_details

Return entity details.

Example arguments:

{ "entity_identifier": "experiment_1", "kg_name": "KANO-ChEBI" }

get_experiment_workflow

Return the full workflow of an experiment.

Example arguments:

{ "experiment_id": "experiment_1" }

Python example (streamable HTTP)

import asyncio
import json
from mcp.client.streamable_http import streamablehttp_client
from mcp.client.session import ClientSession

SERVER_URL = "https://scp.intern-ai.org.cn/api/v1/mcp/37/SciGraph"

async def main():
    transport = streamablehttp_client(
        url=SERVER_URL,
        headers={"SCP-HUB-API-KEY": "sk-xxx"},
    )
    read, write, get_session_id = await transport.__aenter__()

    session_ctx = ClientSession(read, write)
    session = await session_ctx.__aenter__()
    await session.initialize()

    # Example: stats for KANO-ChEBI
    result = await session.call_tool(
        "get_kg_statistics",
        arguments={"kg_name": "KANO-ChEBI"},
    )
    data = json.loads(result.content[0].text)
    print(data)

    await session_ctx.__aexit__(None, None, None)
    await transport.__aexit__(None, None, None)

if __name__ == "__main__":
    asyncio.run(main())

Citation

Fang, Y., Zhang, Q., Zhang, N. et al. (2023). Knowledge graph-enhanced molecular contrastive learning with functional prompt. Nature Machine Intelligence, 5, 542–553. https://doi.org/10.1038/s42256-023-00654-0

Reference

For the full scraped page text, read:

• references/source.md