io.github.cyanheads/ensembl-mcp-server

---

Overview

Gene, sequence, and variant data for vertebrates and other model organisms from the Ensembl REST API. Look up genes, fetch sequences, predict variant consequences, find orthologs, and cross-reference external databases from any MCP client. Runs as a stdio process, a local Streamable HTTP server, or the public hosted endpoint above.

Tools

Tool	Description
ensembl_list_species	List species supported by Ensembl with display name, common name, assembly, taxon ID, and division
ensembl_lookup_gene	Resolve a gene by symbol + species or by stable ID to its Ensembl ID, genomic location, biotype, and transcript list
ensembl_get_sequence	Fetch the DNA, cDNA, CDS, or protein sequence for a gene, transcript, protein, or genomic region
ensembl_query_region	Find genomic features (genes, transcripts, variants, regulatory elements, exons) overlapping a chromosomal region
ensembl_predict_variant	Predict functional consequences of a sequence variant using the Ensembl Variant Effect Predictor (VEP)
ensembl_get_homology	Find orthologs and/or paralogs of a gene across species with percent identity and taxonomy level
ensembl_get_xrefs	Retrieve cross-database references for a gene — HGNC, UniProt, EntrezGene, OMIM, RefSeq, Reactome, and others

Resources

Resource	Description
ensembl://gene/{id}	Gene record by stable ID (ENSG…) — location, biotype, description, and transcript list
ensembl://transcript/{id}	Transcript record by stable ID (ENST…) — parent gene, location, biotype, canonical flag, and length
ensembl://species	Supported Ensembl species for the endpoint default division (vertebrates on the default endpoint)
ensembl://species/{division}	Supported species in one division (EnsemblVertebrates, EnsemblPlants, EnsemblFungi, EnsemblMetazoa, EnsemblProtists)

All resource data is also reachable via the ensembl_list_species tool, which additionally filters by name.

Prompts

Prompt	Description
ensembl_gene_dossier	Structured workflow for assembling a complete gene profile: symbol → ID + location → sequence → variants → orthologs → xrefs

Capability reference

ensembl_list_species _tool

• Filter by division (EnsemblVertebrates, EnsemblPlants, EnsemblFungi, EnsemblMetazoa, EnsemblProtists) or nameContains for a local substring match against name, display name, and common name
• Omit division to return the endpoint default division (vertebrates, ~356 species on the default GRCh38 endpoint)
• Returns internal name (the value every other tool expects), display name, common name, taxon ID, assembly, and division
• Required first step — species names like homo_sapiens are opaque to non-biologists

---

ensembl_lookup_gene _tool

• Exactly one of symbol (+ optional species, default homo_sapiens), id, ids (batch, up to 20), or symbols (batch, up to 20)
• expand_transcripts (default false) adds the full transcript list with biotype and canonical flag
• Batch modes (ids/symbols) return a succeeded/failed split with per-item error strings instead of failing the call
• Errors: not_found, invalid_species, no_input, conflicting_input

---

ensembl_get_sequence _tool

• type: genomic (default, includes introns), cdna (spliced), cds (coding only), protein
• Accepts a stable ID (ENSG…/ENST…/ENSP…) or a region — species:chr:start-end, or bare chr:start-end with species set
• expand_5prime / expand_3prime (default 0) extend flanking base pairs for genomic and region queries
• protein and cds require a transcript or protein ID, not a gene ID
• Every response states length so callers can budget context before consuming large sequences
• Errors: not_found, type_mismatch, missing_species

---

ensembl_query_region _tool

• region in chr:start-end format; feature array defaults to ["gene"], also accepts transcript, variation, regulatory, exon; optional biotype filter
• Defaults to genes only — requesting variation on a large locus can return 44,000+ features
• Exon rows carry a parentId and rank, since one exon is reported once per parent transcript
• Errors: invalid_region, invalid_species

---

ensembl_predict_variant _tool

• variant accepts HGVS (transcript-relative or genomic), region+allele (chr:start:end:strand/allele), or a dbSNP rsID
• max_transcript_consequences (default 10) and max_pubmed_ids_per_variant (default 10) cap large VEP results; set either to 0 for the full set, or include_all_colocated_pubmed: true for uncapped PubMed IDs
• Returns most severe consequence term, per-transcript impact (HIGH/MODERATE/LOW/MODIFIER), and colocated known variants with clinical significance
• Totals (transcriptConsequencesTotal, pubmedTotal) are always reported even when capped
• Errors: invalid_notation, not_found

---

ensembl_get_homology _tool

• Exactly one of symbol (+ species, default homo_sapiens) or id; optional target_species filter
• type: orthologues (default), paralogues, or all
• max_results caps the homolog list (default 25, 0 uncapped); totalCount always reports the true count available
• Errors: not_found, no_input, conflicting_input

---

ensembl_get_xrefs _tool

• id (ENSG…/ENST…) required; optional dbname filter (e.g. HGNC, Uniprot_gn, EntrezGene, MIM_GENE, RefSeq_mRNA, Reactome, GO)
• Uses the xrefs/id endpoint, returning the full cross-reference set (56+ entries for well-annotated genes like BRCA2)
• Errors: not_found

---

ensembl://gene/{id} _resource

• Returns location, biotype, description, and transcript list for a gene stable ID (ENSG…); version suffix optional
• Errors: not_found

---

ensembl://transcript/{id} _resource

• Returns parent gene, location, biotype, canonical flag, and length for a transcript stable ID (ENST…); version suffix optional
• Errors: not_found

---

ensembl://species _resource

• No parameters — returns the endpoint default division (vertebrates, ~356 species on the default GRCh38 endpoint)
• For a named division, read ensembl://species/{division} instead

---

ensembl://species/{division} _resource

• division required: EnsemblVertebrates, EnsemblPlants, EnsemblFungi, EnsemblMetazoa, or EnsemblProtists

---

ensembl_gene_dossier _prompt

• Arguments: gene_symbol required; species optional (default homo_sapiens)
• Sequences a 7-step workflow: resolve the gene → fetch the protein sequence → find variants in the locus → predict variant consequences → find cross-species orthologs → get external database IDs → synthesize the dossier

Features

Built on @cyanheads/mcp-ts-core: stdio and Streamable HTTP transports, pluggable auth (none / jwt / oauth), swappable storage (in-memory, filesystem, Supabase, Cloudflare KV/R2/D1), structured logging with optional OpenTelemetry tracing.

Ensembl-specific:

• Keyless REST API — no API key required; Ensembl REST is fully public at 55,000 req/hr
• Rate-limit-aware service layer: tracks x-ratelimit-remaining, retries 429 with Retry-After, and retries transient 5xx
• Batch POST endpoints used throughout — POST /lookup/id (up to 50 IDs) and POST /lookup/symbol/{species} reduce N+1 round trips in multi-gene workflows
• GRCh37 legacy support via ENSEMBL_BASE_URL — point the entire server at https://grch37.rest.ensembl.org for clinical workflows on the older assembly
• All coordinate-bearing responses echo the assembly name so agents never see a bare genomic position without assembly context

Agent-friendly output:

• Sequence character count stated on every ensembl_get_sequence response so callers can budget context before consuming large genomic sequences
• ensembl_list_species is explicitly the discovery step — tool descriptions call out the opaque internal-name format and direct agents to it before using species-dependent tools
• Cross-tool chaining made explicit: xref IDs from ensembl_get_xrefs are described as inputs for protein and literature servers; the ensembl_gene_dossier prompt sequences all 6 tools into one research workflow

Getting started

Public Hosted Instance

A public instance is available at https://ensembl.caseyjhand.com/mcp — no installation required. Point any MCP client at it via Streamable HTTP:

{
  "mcpServers": {
    "ensembl-mcp-server": {
      "type": "streamable-http",
      "url": "https://ensembl.caseyjhand.com/mcp"
    }
  }
}

Self-Hosted / Local

Add the following to your MCP client configuration file.

{
  "mcpServers": {
    "ensembl-mcp-server": {
      "type": "stdio",
      "command": "bunx",
      "args": ["@cyanheads/ensembl-mcp-server@latest"],
      "env": {
        "MCP_TRANSPORT_TYPE": "stdio",
        "MCP_LOG_LEVEL": "info"
      }
    }
  }
}

Or with npx (no Bun required):

{
  "mcpServers": {
    "ensembl-mcp-server": {
      "type": "stdio",
      "command": "npx",
      "args": ["-y", "@cyanheads/ensembl-mcp-server@latest"],
      "env": {
        "MCP_TRANSPORT_TYPE": "stdio",
        "MCP_LOG_LEVEL": "info"
      }
    }
  }
}

Or with Docker:

{
  "mcpServers": {
    "ensembl-mcp-server": {
      "type": "stdio",
      "command": "docker",
      "args": [
        "run", "-i", "--rm",
        "-e", "MCP_TRANSPORT_TYPE=stdio",
        "ghcr.io/cyanheads/ensembl-mcp-server:latest"
      ]
    }
  }
}

For Streamable HTTP, set the transport and start the server:

MCP_TRANSPORT_TYPE=http MCP_HTTP_PORT=3010 bun run start:http
# Server listens at http://localhost:3010/mcp

Prerequisites

• Bun v1.4.0 or higher (or Node.js v24+).
• No API key required — Ensembl REST is fully public.

Installation

1. Clone the repository:

git clone https://github.com/cyanheads/ensembl-mcp-server.git

2. Navigate into the directory:

cd ensembl-mcp-server

3. Install dependencies:

bun install

4. Configure environment:

cp .env.example .env
# edit .env if you need to override ENSEMBL_BASE_URL (e.g. for GRCh37)

Configuration

All configuration is validated at startup via Zod schemas in src/config/server-config.ts.

Variable	Description	Default
ENSEMBL_BASE_URL	Ensembl REST API base URL. Override for GRCh37 (https://grch37.rest.ensembl.org) or a local mirror.	https://rest.ensembl.org
MCP_TRANSPORT_TYPE	Transport: stdio or http	stdio
MCP_HTTP_PORT	HTTP server port	3010
MCP_HTTP_ENDPOINT_PATH	HTTP endpoint path	/mcp
MCP_SESSION_MODE	HTTP session mode: auto, stateful, or stateless. Schema default auto resolves to stateful; this server explicitly uses stateless.	stateless
MCP_AUTH_MODE	Authentication: none, jwt, or oauth	none
MCP_LOG_LEVEL	Log level (debug, info, warning, error, etc.)	info
LOGS_DIR	Directory for log files (Node.js only)	<project-root>/logs
OTEL_ENABLED	Enable OpenTelemetry	false

See .env.example for the full list of optional overrides.

Running the server

Local development

• Build and run:

  # One-time build
  bun run rebuild
  
  # Run the built server
  bun run start:stdio
  # or
  bun run start:http
  

• Run checks and tests:

  bun run devcheck   # Lint, format, typecheck, security
  bun run test       # Vitest test suite
  bun run lint:mcp   # Validate MCP definitions against spec
  

Docker

docker build -t ensembl-mcp-server .
docker run --rm -p 3010:3010 ensembl-mcp-server

The Dockerfile defaults to HTTP transport, stateless session mode, and logs to /var/log/ensembl-mcp-server. OpenTelemetry peer dependencies are installed by default — build with --build-arg OTEL_ENABLED=false to omit them.

Project structure

Directory	Purpose
src/index.ts	createApp() entry point — registers tools/resources/prompts and inits services
src/config	Server-specific environment variable parsing and validation with Zod
src/mcp-server/tools	Tool definitions (*.tool.ts) — 7 tools
src/mcp-server/resources	Resource definitions (*.resource.ts) — gene, transcript, species
src/mcp-server/prompts	Prompt definitions (*.prompt.ts) — gene dossier workflow
src/services/ensembl	Ensembl REST API client — HTTP, rate-limit handling, retry, error normalization
tests/	Unit and integration tests mirroring src/

Development guide

See CLAUDE.md for development guidelines and architectural rules. The short version:

• Handlers throw, framework catches — no try/catch in tool logic
• Use ctx.log for request-scoped logging, ctx.state for tenant-scoped storage
• Register new tools and resources in the createApp() arrays in src/index.ts
• Wrap external API calls: validate raw → normalize to domain type → return output schema; never fabricate missing fields

Contributing

Issues are welcome. Run checks and tests before submitting:

bun run devcheck
bun run test

License

Apache-2.0 — see LICENSE for details.