kegg-mcp
An MCP server that exposes the KEGG REST API — pathways, genes, compounds, reactions, enzymes, diseases, drugs, modules, orthology, glycans, and BRITE hierarchies, with cross-database linking. Provides 34 tools and 8 resource templates.
README
kegg-mcp
An MCP server that exposes the KEGG REST API — pathways, genes, compounds, reactions, enzymes, diseases, drugs, modules, orthology, glycans, and BRITE hierarchies, with cross-database linking. 34 tools, 8 resource templates, runs over stdio.
Install
npm install -g kegg-mcp
…or run it once with no install:
npx -y kegg-mcp
Requires Node.js ≥ 20.
Connect your MCP client
Add a kegg server to your client config (Claude Desktop, Cursor, VS Code, Cline, …):
{
"mcpServers": {
"kegg": {
"command": "npx",
"args": ["-y", "kegg-mcp"]
}
}
}
Prefer a global install? Use "command": "kegg-mcp" with no args.
Highlights
- 34 tools across every major KEGG database
- 8 resource templates (
kegg://pathway/{id},kegg://gene/{org}:{id}, …) - Polite rate limiting (≤ 3 req/s), retries with exponential backoff, 30 s timeouts, keep-alive
- Native
fetch— zero runtime deps beyond the MCP SDK - TypeScript, unit tests with real KEGG fixtures, live integration tests, CI on Node 20/22/24
Quick examples
search_pathways(query="glycolysis")
get_pathway_info(pathway_id="hsa00010")
get_pathway_genes(pathway_id="hsa00010")
search_genes(query="insulin", organism_code="hsa")
get_gene_info(gene_id="hsa:3630", include_sequences=true)
search_compounds(query="glucose")
get_compound_info(compound_id="C00031")
convert_identifiers(source_db="hsa", target_db="ncbi-geneid")
batch_entry_lookup(entries=["hsa:3630", "hsa:3631"])
kegg_health_check()
Tools (34)
Info — get_database_info, list_organisms
Pathways — search_pathways, get_pathway_info, get_pathway_genes, get_pathway_compounds, get_pathway_reactions
Genes — search_genes, get_gene_info, get_gene_orthologs
Compounds — search_compounds, get_compound_info, get_compound_reactions
Reactions & enzymes — search_reactions, get_reaction_info, search_enzymes, get_enzyme_info
Disease & drugs — search_diseases, get_disease_info, search_drugs, get_drug_info, get_drug_interactions
Modules & orthology — search_modules, get_module_info, search_ko_entries, get_ko_info
Glycans & BRITE — search_glycans, get_glycan_info, search_brite, get_brite_info
Cross-db & batch — convert_identifiers, find_related_entries, batch_entry_lookup
Health — kegg_health_check
Resource templates (8)
kegg://pathway/{pathway_id} · kegg://gene/{org}:{gene_id} · kegg://compound/{compound_id} · kegg://reaction/{reaction_id} · kegg://disease/{disease_id} · kegg://drug/{drug_id} · kegg://organism/{org_code} · kegg://search/{database}/{query}
Develop
git clone https://github.com/AriaShishegaran/kegg-mcp.git
cd kegg-mcp
npm install
npm run build # outputs to dist/
npm test # unit tests
npm run test:integration # live tests against rest.kegg.jp
| Script | What it does |
|---|---|
npm run build |
Compile to dist/ |
npm run typecheck |
Type-check only |
npm run lint |
ESLint |
npm test |
Unit tests |
npm run test:integration |
Live tests against rest.kegg.jp |
npm start |
Run the server |
Releases are automated via release-please from Conventional Commits. See CONTRIBUTING.md.
KEGG API terms
rest.kegg.jp is academic use only. Commercial use requires a license from Kanehisa Laboratories.
License
MIT © Aria
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。