Gget

You can access powerful bioinformatics tools through a dedicated MCP server that exposes gget capabilities. This server provides a structured, AI-friendly interface to query gene and sequence data, perform sequence analysis, expression and enrichment analyses, protein structure predictions, and cancer/mutation insights, all via a standard MCP connection.

How to use

Connect to the gget MCP server using an MCP client. The server exposes tools for gene search, sequence retrieval, alignment, enrichment analysis, protein structure prediction, and more. Use natural language prompts to request data, analyses, or file outputs, and let the MCP client handle the orchestration of the underlying gget functions.

How to install

Prerequisites: you need a Unix-like shell, and you should have either uv or Node.js installed depending on your chosen workflow.

# Quick start: install uv (optional, uvx can auto-install)
curl -LsSf https://astral.sh/uv/install.sh | sh

# Verify installation
uv --version
uvx --version

Run the MCP server in STDIO mode using uvx. This starts a local server you can connect to from MCP clients.

uvx gget-mcp stdio

If you prefer HTTP mode to interact over a web server, run the server in HTTP mode.

uvx gget-mcp http

If you want Server-Sent Events (SSE) mode for real-time communication, run the server in SSE mode.

uvx gget-mcp sse

Additional configuration and notes

The server can be configured for different MCP client setups. You can set up STDIO mode to write outputs to disk or HTTP mode to serve a web interface. The recommended STDIO configuration uses uvx to run the server from the latest gget-mcp package in stdio mode.

{
  "mcpServers": {
    "gget-mcp": {
      "command": "uvx",
      "args": ["--from", "gget-mcp@latest", "stdio"]
    }
  }
}

Troubleshooting and tips

If you encounter issues with uvx, try cleaning the uv cache and re-running the server. Common steps include ensuring uv and node are in your PATH and restarting the client if you see a timeout on first launch.

For per-session debugging, inspect the running MCP server’s status in your MCP client or device logs. If a request times out or an operation stalls, verify your network connection and try restarting the server mode you are using (stdio, http, or sse).

Tools and capabilities you can access

The server exposes a broad set of bioinformatics tools via MCP. Here are the core functions you can invoke through your MCP client.

- gget_search: Find Ensembl IDs associated with search terms
- gget_info: Fetch detailed information for Ensembl IDs
- gget_seq: Retrieve nucleotide or amino acid sequences
- gget_ref: Get reference genome information from Ensembl
- gget_blast: BLAST nucleotide or amino acid sequences
- gget_blat: Find genomic locations of sequences
- gget_muscle: Align multiple sequences
- gget_archs4: Get gene expression data from ARCHS4
- gget_enrichr: Perform gene set enrichment analysis
- gget_pdb: Fetch protein structure data from PDB
- gget_alphafold: Predict protein structure
- gget_cosmic: Search COSMIC cancer mutations
- gget_cellxgene: Query single-cell RNA-seq data

Available tools

gget_search

Find Ensembl IDs associated with search terms

gget_info

Fetch detailed information for Ensembl IDs

gget_seq

Retrieve nucleotide or amino acid sequences

gget_ref

Get reference genome information from Ensembl

gget_blast

BLAST nucleotide or amino acid sequences

gget_blat

Find genomic locations of sequences

gget_muscle

Align multiple sequences

gget_archs4

Get gene expression data from ARCHS4

gget_enrichr

Perform gene set enrichment analysis

gget_pdb

Fetch protein structure data from PDB

gget_alphafold

Predict protein structure using AlphaFold

gget_cosmic

Search COSMIC database for cancer mutations

gget_cellxgene

Query single-cell RNA-seq data from CellxGene