AlphaGenome

The AlphaGenome MCP Server exposes a production-grade interface to Google DeepMind's AlphaGenome API, enabling researchers to run advanced genomic analyses through natural language commands and structured tool calls. It supports core predictions (DNA sequence analysis, genomic interval analysis, variant effect prediction, and variant scoring), batch processing for sequences, intervals, and variants, and advanced scoring and analysis workflows. Users can invoke individual tools (e.g., predict_dna_sequence, predict_genomic_interval, predict_variant_effect, score_variant) or batch tools (predict_sequences, predict_intervals, predict_variants, score_variants) as MCP tools, and combine these with auxiliary utilities for validation, parsing, and metadata retrieval. The server is designed for production use with real API integration, robust error handling, and parallel processing where applicable.

Prerequisites:

• Docker 20.10+ and Docker Compose 2.0+ (for containerized deployment)
• AlphaGenome API key from Google DeepMind
• Network access to pull required images and reach the AlphaGenome API

Installation steps:

1. Clone the repository:

git clone https://github.com/Augmented-Nature/AlphaGenome-MCP-Server.git
cd AlphaGenome-MCP-Server

2. Obtain and set your AlphaGenome API key:

# Example: store key in an environment file or directly set in a container environment
echo "ALPHAGENOME_API_KEY=your-api-key-here" > .env

3. Build/run with Docker Compose (recommended for isolation and reproducibility):

# Ensure environment variables are loaded inside the container environment
docker-compose up -d

4. Confirm the MCP server is running and accessible. You should see the Alphagenome MCP server container active and the index.js process running inside it. If needed, inspect logs:

docker-compose logs -f

5. Add MCP configuration to your MCP client by referencing the server name and command pattern shown in the example below. The server uses a Docker-based Node.js entry point inside the container; ensure the container name matches alphagenome-mcp-server if you customize deployment.

{
  "mcpServers": {
    "alphagenome-server": {
      "type": "stdio",
      "command": "docker",
      "args": [
        "exec",
        "-i",
        "alphagenome-mcp-server",
        "node",
        "/app/build/index.js"
      ],
      "disabled": false,
      "autoApprove": [],
      "timeout": 30
    }
  }
}

Tips and common issues:

• Keep your ALPHAGENOME_API_KEY secure; rotate keys as recommended by your organization and do not commit keys to version control.
• Ensure the Docker container alphagenome-mcp-server is running before starting your MCP client; use docker-compose start to initialize services.
• If you encounter API constraint messages (e.g., interval width or batch size limits), adjust max_workers or interval parameters in your tool calls.
• The server supports both single-tool calls and batch processing; for high-throughput workloads, enable parallel processing via max_workers settings in batch tool arguments.
• Monitor logs regularly during initial setup to catch misconfigurations or network/auth failures early.