Ensembl MCP Server
Used as the runtime environment for the MCP server, enabling the execution of TypeScript/JavaScript code for genomic data querying
Used as the programming language for implementing the MCP server, providing type safety for genomic data structures and API interactions
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@Ensembl MCP Server find genes overlapping chr17:43000000-44000000 in human"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
🧬 Ensembl API MCP Server
A full-featured Model Context Protocol (MCP) server that exposes Ensembl’s REST API. Built using the TypeScript MCP SDK
Comprehensive coverage – 10 tools map to functional areas instead of 100 + individual endpoints, yet still expose nearly the whole API.
Production-ready – TypeScript throughout, robust error handling, and a tidy API-client layer.
Biologist-friendly – grouped by biological task (genes, variants, compara…), not by low-level REST paths.
Listed on:
Related MCP server: GeneOntology MCP Server
Use cases:
🧬 Gene information – fetch details by ID or symbol
🔍 Gene search – scan genes across any species
🧬 Sequence retrieval – pull DNA for any genomic region
🔬 Variant data – explore variants and their annotations
📊 Transcript info – inspect transcripts and isoforms
🌍 Multi-species – every species in Ensembl, right here
🔗 Cross-references – hop to external databases in one call
⚡ Rate-limited – built-in throttling keeps you within Ensembl limits
Installation
Choose your preferred installation method:
Option 1: Via Smithery
Visit Smithery - Ensembl MCP Server. The most common platform options include:
# claude code:
npx -y @smithery/cli@latest install @effieklimi/ensembl-mcp-server --client claude --key your-smithery-secret-key
# cursor:
npx -y @smithery/cli@latest install @effieklimi/ensembl-mcp-server --client cursor --key your-smithery-secret-key
# vscode:
npx -y @smithery/cli@latest install @effieklimi/ensembl-mcp-server --client vscode --key your-smithery-secret-key
# windsurf:
npx -y @smithery/cli@latest install @effieklimi/ensembl-mcp-server --client windsurf --key your-smithery-secret-keyCheck the MCP's smithery link for additional platform options.
Option 2: Local Development Setup
For development or custom setups:
Clone and install dependencies:
git clone https://github.com/effieklimi/ensembl-mcp-server.git cd ensembl-mcp-server npm installConfigure Claude Desktop manually:
Edit your config file:
macOS:
~/Library/Application Support/Claude/claude_desktop_config.jsonWindows:
%APPDATA%/Claude/claude_desktop_config.json
Add this server configuration:
{ "mcpServers": { "ensembl": { "command": "npm", "args": ["run", "start"], "cwd": "/absolute/path/to/ensembl-mcp-server" } } }Restart Claude Desktop - The Ensembl tools will appear in your available tools
Development Setup
# Development with hot reload
npm run dev
# Run tests
npm test
# Production build (optional)
npm run build
npm run start:prodContributing
We'd love your help! Here's how to get started:
Quick Contact
Email the dev: effie@effie.bio
Development Workflow
Fork the repository
Clone your fork:
git clone https://github.com/YOUR_USERNAME/ensembl-mcp-server.git cd ensembl-mcp-serverInstall dependencies:
npm installRun tests to make sure everything works:
npm testStart development server:
npm run devMake your changes and test thoroughly
Submit a pull request
Available Scripts
npm run dev- Development with hot reloadnpm run start- Run the servernpm test- Run all testsnpm run build- Compile TypeScript (optional)npm run start:prod- Run compiled version
The ten tools (with endpoints)
1 · ensembl_feature_overlap
Find genes, transcripts, or regulatory elements that overlap a region or another feature.
GET /overlap/region/:species/:region
GET /overlap/id/:idTypical asks: “Which genes sit in chr17:43-44 Mb?” – “What overlaps BRCA1?”
2 · ensembl_regulatory
Regulatory features, binding matrices and related annotations.
GET /overlap/region/:species/:region (with regulatory filters)
GET /overlap/translation/:id (regulatory features on proteins)
GET /species/:species/binding_matrix/:binding_matrix_stable_idUse cases: TF-binding sites, regulatory annotation.
3 · ensembl_protein_features
Protein-level domains and functional sites.
GET /overlap/translation/:idUse cases: protein domains, signal peptides, catalytic residues.
4 · ensembl_meta
Server metadata, species lists, release info, and diagnostics.
GET /info/ping
GET /info/rest
GET /info/software
GET /info/data
GET /info/species
GET /info/divisions
GET /info/assembly/:species
GET /info/biotypes/:species
GET /info/analysis/:species
GET /info/external_dbs/:species
GET /info/variation/:species
GET /archive/id/:id
POST /archive/idTypical asks: “Which assemblies do you have for human?” – server health checks.
5 · ensembl_lookup
Translate IDs ↔ symbols, pull xrefs, recode variants.
GET /lookup/id/:id
GET /lookup/symbol/:species/:symbol
POST /lookup/id
POST /lookup/symbol
GET /xrefs/id/:id
GET /xrefs/symbol/:species/:symbol
GET /xrefs/name/:species/:name
GET /variant_recoder/:species/:id
POST /variant_recoder/:speciesUse cases: “What is BRCA1’s Ensembl ID?” – cross-reference UniProt.
6 · ensembl_sequence
Retrieve DNA, RNA or protein sequences.
GET /sequence/id/:id
GET /sequence/region/:species/:region
POST /sequence/id
POST /sequence/regionUse cases: gene FASTA, transcript cDNA, genomic regions.
7 · ensembl_mapping
Coordinate conversion (genome ↔ cDNA/CDS/protein) and assembly lift-over.
GET /map/cdna/:id/:region
GET /map/cds/:id/:region
GET /map/translation/:id/:region
GET /map/:species/:asm_one/:region/:asm_twoUse cases: map CDS to GRCh38, convert protein to genome coords.
8 · ensembl_compara
Comparative genomics—homology, gene trees, alignments.
GET /homology/id/:species/:id
GET /homology/symbol/:species/:symbol
GET /genetree/id/:id
GET /genetree/member/symbol/:species/:symbol
GET /genetree/member/id/:species/:id
GET /cafe/genetree/id/:id
GET /cafe/genetree/member/symbol/:species/:symbol
GET /cafe/genetree/member/id/:species/:id
GET /alignment/region/:species/:regionUse cases: find orthologs, build phylogenies, pull species alignments.
9 · ensembl_variation
Variant lookup, VEP consequences, LD, phenotype mapping.
GET /variation/:species/:id
GET /variation/:species/pmcid/:pmcid
GET /variation/:species/pmid/:pmid
POST /variation/:species
GET /vep/:species/hgvs/:hgvs_notation
POST /vep/:species/hgvs
GET /vep/:species/id/:id
POST /vep/:species/id
GET /vep/:species/region/:region/:allele
POST /vep/:species/region
GET /ld/:species/:id/:population_name
GET /phenotype/variant/:species/:id
GET /phenotype/region/:species/:region
GET /transcript_haplotypes/:species/:idUse cases: VEP predictions, LD blocks, phenotype associations.
10 · ensembl_ontotax
Ontology term search and NCBI taxonomy traversal.
GET /ontology/id/:id
GET /ontology/name/:name
GET /taxonomy/id/:id
GET /taxonomy/name/:nameUse cases: GO term look-up, phenotype ontology, taxonomic classification.
Installing via Smithery
To install ensembl-mcp-server for Claude Desktop automatically via Smithery:
npx -y @smithery/cli install @effieklimi/ensembl-mcp-server --client claudeAvailable Tools
10 toolsensembl_comparaC
Comparative genomics: gene trees, homology, species alignments, and evolutionary analysis. Covers /genetree/, /homology/, /alignment/* endpoints.
Inputs: gene_id, gene_symbol, region, analysis_type, +4 more.
ensembl_feature_overlapB
Find genomic features (genes, transcripts, regulatory elements) that overlap with a genomic region or specific feature. Automatically handles assembly-specific format variations (GRCh38/hg38, chromosome naming conventions, coordinate systems). Covers /overlap/region and /overlap/id endpoints.
Inputs: region, feature_id, species, feature_types, +1 more.
ensembl_lookupB
Look up genes, transcripts, variants by ID or symbol. Get cross-references and perform ID translation. Covers /lookup/* and /xrefs/* endpoints plus variant_recoder.
Inputs: identifier (required), lookup_type, species, expand, +1 more.
ensembl_mappingC
Map coordinates between different coordinate systems (genomic ↔ cDNA/CDS/protein) and between genome assemblies. Covers /map/* endpoints.
Inputs: coordinates (required), mapping_type (required), feature_id, source_assembly, +2 more.
ensembl_metaA
Get server metadata, data releases, species info, and system status. Covers /info/* endpoints and /archive/id for version tracking.
Inputs: info_type, species, archive_id, division.
ensembl_ontotaxC
Ontology term search and NCBI taxonomy traversal. Search GO terms, phenotype ontologies, and taxonomic classifications.
Inputs: term, ontology, term_id, species, +1 more.
ensembl_protein_featuresC
Get protein-level features, domains, and annotations for proteins and translations.
Inputs: protein_id (required), feature_type, species.
ensembl_regulatoryC
Get regulatory features, binding matrices, and regulatory annotations. Covers regulatory overlap endpoints and binding matrix data.
Inputs: region, protein_id, binding_matrix_id, species, +1 more.
ensembl_sequenceC
Retrieve DNA, RNA, or protein sequences for genes, transcripts, regions. Covers /sequence/id and /sequence/region endpoints.
Inputs: identifier (required), sequence_type, species, format, +1 more.
ensembl_variationC
Variant analysis: VEP consequence prediction, variant lookup, LD analysis, phenotype mapping, haplotypes. Covers /variation/, /vep/, /ld/, /phenotype/ endpoints.
Inputs: variant_id, region, hgvs_notation, analysis_type, +4 more.
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
TDQS
Scored across 10 tools
Each tool has a clearly distinct purpose targeting specific genomic data types or operations, such as comparative genomics, feature overlap, sequence retrieval, or variant analysis. The descriptions explicitly differentiate their scopes (e.g., 'ensembl_compara' for evolutionary analysis vs. 'ensembl_sequence' for sequence retrieval), with no apparent overlap that could cause agent misselection.
All tool names follow a consistent 'ensembl_' prefix with descriptive suffixes (e.g., 'compara', 'lookup', 'mapping'), using snake_case uniformly throughout. This predictable pattern enhances readability and helps agents quickly identify the tool's domain without confusion.
With 10 tools, the server is well-scoped for genomic data access and analysis, covering key areas like sequence retrieval, variant analysis, and metadata. Each tool earns its place by addressing distinct aspects of the Ensembl database, avoiding bloat while providing comprehensive coverage for typical bioinformatics workflows.
The tool set offers complete coverage for genomic data operations, including lookup, mapping, sequence retrieval, variation analysis, and metadata access. It supports full CRUD-like workflows (e.g., from data query to analysis) with no obvious gaps, ensuring agents can handle end-to-end tasks without dead ends.
Maintenance
Related MCP Connectors
A comprehensive Model Context Protocol (MCP) server that enables AI assistants to interact with yo…
A Model Context Protocol server for Wix AI tools
Provide detailed Pokémon data and information through a standardized MCP interface. Enable LLMs an…
Related MCP Servers
- AlicenseAqualityDmaintenanceA comprehensive Model Context Protocol (MCP) server that provides access to the Ensembl REST API for genomic data, comparative genomics, and biological annotations.194MIT
- AlicenseBqualityDmaintenanceA comprehensive Model Context Protocol server for accessing Gene Ontology (GO) data, enabling AI systems to perform ontology-based analysis, gene annotation research, and functional enrichment studies.47MIT
- AlicenseNot gradedqualityDmaintenanceA Model Context Protocol server that enables retrieval of -omics data from sources like PDB, ChEMBL, UniProt, PubChem, and HUGO.5GPL 3.0
- AlicenseNot gradedqualityCmaintenanceEnables querying Ensembl genomic data including gene lookup, sequence retrieval, homology, variation, and variant effect prediction via MCP tools.96 npmMIT