RNAcentral MCP Server
OfficialClick 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., "@RNAcentral MCP Server find overlapping ncRNAs for BRCA1 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.
RNAcentral MCP Server
This is an MCP server that provides a comprehensive interface to the RNAcentral database, allowing for complex searches, sequence mapping, genomic analysis, and metadata retrieval for non-coding RNA sequences.
📖 Full documentation: https://rnacentral.github.io/rnacentral-mcp-server/ — installation guides for Claude Desktop / Claude Code / other MCP clients, per-tool reference, and examples.
Features
Comprehensive Search: Query RNAcentral using natural language or filters (RNA type, taxon, expert database). Combines EBI Search and RNAcentral API data for enriched metadata (Rfam hits, GO annotations).
Sequence Search: Search for RNA sequences across multiple databases to find identical or similar entries.
Bidirectional ID Mapping: Map between RNAcentral URS IDs and external database identifiers (miRBase, Ensembl, HGNC, etc.) with automatic taxonomy resolution.
Bulk Sequence Export: Export search results in FASTA or Parquet formats, ideal for downstream analysis or machine learning datasets.
Genomic Overlap (Ensembl): Find non-coding RNAs overlapping specific genomic coordinates or gene symbols using the Ensembl REST API.
2D Structure Diagrams: Retrieve secondary structure (2D) diagrams in SVG format for RNAs with known or predicted folds.
Literature Summaries: Access AI-generated literature summaries for RNA sequences to understand their biological context.
Related MCP server: nucleotide_archive_mcp
Prerequisites
Python 3.14 or higher
uv (recommended) or pip
Installation
The server is published on PyPI as rnacentral-mcp-server. The easiest way to run it is with uvx, which fetches and runs it without a manual install:
uvx --from rnacentral-mcp-server run-serverOr install it into an environment with pip / uv:
pip install rnacentral-mcp-server
# or
uv add rnacentral-mcp-serverThis makes the run-server console script available.
From source (for development)
Set up a Python environment:
# Using uv (recommended)
uv venv
source .venv/bin/activate # On Windows: .venv\Scripts\activate
# Add dependencies
uv add "mcp[cli]" aiohttpInstall the package in editable mode:
uv pip install -e .Running the Server
Development Mode (with MCP Inspector)
mcp dev rnacentral_sequence_search/server.pyThis will start the server and open the MCP Inspector, allowing you to test the server interactively.
Using with Claude Desktop
To install the server in Claude Desktop, add this to your claude_desktop_config.json. You can optionally specify a --log-dir to save logs to a specific directory:
{
"mcpServers": {
"rnacentral": {
"command": "uvx",
"args": [
"--from",
"rnacentral-mcp-server",
"run-server",
"--log-dir",
"/Users/YOUR_USERNAME/logs/rnacentral"
]
}
}
}(To run the latest unreleased code instead, replace rnacentral-mcp-server with git+https://github.com/rnacentral/rnacentral-mcp-server.git.)
Direct Execution
You can run the server directly using Python. Use the --log-dir argument to specify where to save log files:
python rnacentral_sequence_search/server.py --log-dir ./logsOr using the installed script:
run-server --log-dir ./logsRunning Tests
Run the unit test suite with uv:
uv run pytestUsage Examples
Once the server is running, you can interact with various tools:
Bidirectional ID Mapping
Map an external ID to RNAcentral and see all cross-references:
Tool: map_rna_id
Arguments:
{
"identifier": "MIMAT0000062",
"taxon": "Homo sapiens"
}Genomic Overlap
Find ncRNAs overlapping a specific gene in human:
Tool: get_overlapping_ncrnas
Arguments:
{
"species": "human",
"gene_symbol": "HOTAIR"
}2D Structure Retrieval
Get the secondary structure diagram for a specific URS ID:
Tool: get_secondary_structure_svg
Arguments:
{
"urs_id": "URS0000049E57"
}Bulk Sequence Export
Export a set of sequences matching a search query for machine learning:
Tool: export_sequences
Arguments:
{
"query": "lncRNA",
"taxon": "9606",
"format": "parquet",
"max_length": 500
}Literature Summaries
Get a summary of the known biological role of an RNA:
Tool: get_rna_description
Arguments:
{
"rna_id": "mmu-mir-191"
}Notes
The server handles complex queries by orchestrating multiple upstream APIs (EBI Search, RNAcentral, Ensembl).
Sequence searches poll for results with a timeout to handle varying server loads.
The 2D structure SVG can be used for direct visualization in supporting clients.
For very large exports, use the
export_sequencestool which uses specialized microservices for efficiency.
Available Tools
8 toolsexport_sequencesA
Export bulk RNA sequences matching a query in FASTA or Parquet format. Use this for large-scale data retrieval (e.g., ML dataset generation).
Args: query: Search term (e.g., "telomerase", "hsa-mir-126", "HOTAIR") rna_type: Filter by RNA type (e.g., "miRNA", "lncRNA", "tRNA") taxon: Taxon name or NCBI Taxonomy ID. expert_db: Source database (e.g., "miRBase", "GENCODE", "Ensembl") has_secondary_structure: Filter for RNAs with known 2D structure min_length: Minimum sequence length (nt) max_length: Maximum sequence length (nt) format: Output format ("fasta" or "parquet", default "fasta")
Inputs: query (required), taxon, format, rna_type, +4 more. Returns: result
get_overlapping_ncrnasA
Get non-coding RNAs overlapping a genomic region or a specific gene.
You must provide EITHER:
chromosome, start, and end (e.g., chromosome="1", start=100000, end=200000)
OR gene_symbol (e.g., "BRCA2"), which will be automatically resolved to coordinates via Ensembl.
Args: species: Common name (e.g. "human", "mouse") or scientific name (e.g. "homo_sapiens"). chromosome: Chromosome name (e.g., "1", "X", "chr1"). start: Start coordinate. end: End coordinate. gene_symbol: Gene symbol to lookup coordinates for.
Inputs: species (required), end, start, chromosome, +1 more. Returns: result
get_rna_descriptionB
Get a literature summary description of an RNA sequence from RNAcentral.
Args: rna_id: The ID of the RNA (e.g., "mmu-mir-191")
Returns: A textual description of the RNA sequence based on literature analysis.
Inputs: rna_id (required). Returns: result
get_secondary_structure_svgB
Get the secondary structure (2D diagram) for an RNA sequence in SVG format.
Args: urs_id: The RNAcentral URS ID (e.g., 'URS0000049E57')
Inputs: urs_id (required).
get_sequence_search_resultsA
Check a sequence search job started by search_sequence. Returns progress while the job is running, or a markdown summary of hits once finished. If still running, wait (e.g. 30-60s) and call again.
Args: job_id: The job ID returned by search_sequence
Inputs: job_id (required). Returns: result
map_rna_idA
Bidirectional ID mapping for RNA sequences. Maps between RNAcentral URS IDs and external database identifiers.
Args: identifier: RNAcentral URS ID (e.g., "URS0000759B6D") or External ID (e.g., "MIMAT0000062") taxon: Optional taxon name or NCBI Taxonomy ID. Prefer scientific names (e.g. "Homo sapiens", "Mus musculus") if available.
Inputs: identifier (required), taxon. Returns: result
query_rnacentralB
Comprehensive search for RNA sequences using EBI Search and RNAcentral APIs. Combines text search filters with detailed metadata (Rfam, GO, 2D structure).
Args: query: Search term (e.g., "telomerase", "hsa-mir-126", "HOTAIR") rna_type: Filter by RNA type (e.g., "miRNA", "lncRNA", "tRNA") taxon: Taxon name or NCBI Taxonomy ID. Prefer scientific names (e.g. "Homo sapiens", "Mus musculus") over common names if available. expert_db: Source database (e.g., "miRBase", "GENCODE", "Ensembl") has_secondary_structure: Filter for RNAs with known 2D structure limit: Number of results to return (default 10, max 20)
Inputs: query (required), limit, taxon, rna_type, +2 more. Returns: result
search_sequenceA
Start a background search of the RNAcentral database for a given RNA sequence. Returns a job ID immediately; the search can take minutes to hours under load. Poll get_sequence_search_results with the job ID until it reports results.
Args: sequence: The RNA sequence to search for databases: Optional list of specific databases to search within (default: all databases)
Returns: A job ID to pass to get_sequence_search_results
Inputs: sequence (required), databases. Returns: result
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
TDQS
Scored across 8 tools
Most tools have clear distinct purposes: text/filter search (query_rnacentral), exact sequence search (search_sequence), bulk export (export_sequences), ID mapping, structure retrieval, genomic overlap, and description. The async pair search_sequence/get_sequence_search_results is explicitly linked via job_id documentation. Minor potential confusion between query_rnacentral and export_sequences (both take identical filter params) and between text search vs sequence search, but descriptions clarify them.
Predominant verb_noun pattern (get_secondary_structure_svg, map_rna_id, search_sequence, export_sequences, getoverlapping_ncrnas, get_rna_description). Slight deviations like query_rnacentral and the paired get_sequence_search_results break pure uniformity, but the overall convention is readable and predictable.
Eight tools is well-scoped for an RNA database access server: search, filter query, bulk export, ID mapping, structure, genomic overlap, description, and async polling. Each tool clearly earns its place with no filler or redundant surface.
Coverage across the RNA-lookup lifecycle is strong: sequence search, filtered text query, bulk export, ID mapping, 2D structure, genomic overlap, and descriptions. A notable minor gap is a direct 'fetch full record by URS ID' metadata tool; get_rna_description returns only literature summaries, so agents may rely on query_rnacentral to retrieve record details.
Maintenance
Related MCP Connectors
Search InterPro entries, map domains onto proteins and PDB structures, and scan raw sequences.
Protein research over UniProtKB: search by function, fetch curated records, map IDs, proteomes.
European Nucleotide Archive (ENA) — EMBL-EBI's public record of raw sequencing runs, assemblies…
Query STRING interactions, enrichment, annotations, homology, and PPI networks.
Related MCP Servers
- AlicenseNot gradedqualityFmaintenanceExecutes SPARQL queries against biological and biomedical RDF databases from the RDF Portal, with additional REST API integrations for bioinformatics resources.1MIT
- AlicenseAqualityDmaintenanceAn MCP server for searching and accessing RNA sequencing datasets from the European Nucleotide Archive (ENA), supporting bulk, single-cell, and spatial transcriptomics with advanced filtering and download capabilities.111,224 PyPI1Apache 2.0

io.github.rcsb/rcsb-mcpofficial
AlicenseAqualityBmaintenanceEnables discovery, inspection, and cross-referencing of Protein Data Bank structures through three RCSB APIs: search, data, and sequence coordinates.3612MIT
@pipeworx/enaofficial
AlicenseNot gradedqualityCmaintenanceEnables searching ENA's public sequencing data by organism, study, platform, country or date, retrieving FASTQ/BAM download URLs and checksums, and discovering available result types and fields.91 npmMIT