bio-mcp/bio-mcp-interpro

bio-mcp-interpro

MCP (Model Context Protocol) server for InterPro protein domain and family analysis.

Overview

This MCP server provides access to InterProScan functionality, allowing AI assistants to perform protein domain and family analysis, functional annotation, and GO term mapping.

Features

• Protein domain and family identification
• GO term annotation and pathway mapping
• Support for multiple output formats (TSV, XML, JSON, GFF3)
• Background job processing for large datasets
• File size limits and timeout protection
• Temporary file management
• Async execution with proper error handling

Installation

Using pip

pip install bio-mcp-interpro

From source

git clone https://github.com/bio-mcp/bio-mcp-interpro
cd bio-mcp-interpro
pip install -e .

Configuration

Configure your MCP client (e.g., Claude Desktop) by adding to your configuration:

{
  "mcp-servers": {
    "bio-interpro": {
      "command": "python",
      "args": ["-m", "src.server_with_queue"]
    }
  }
}

Environment Variables

• BIO_MCP_MAX_FILE_SIZE: Maximum input file size (default: 100MB)
• BIO_MCP_TIMEOUT: Command timeout in seconds (default: 1800)
• BIO_MCP_INTERPRO_PATH: Path to InterProScan executable (default: interproscan.sh)

Usage

Once configured, the AI assistant can use the following tools:

interpro_run_async

Run InterProScan protein domain and family analysis (background job)

Parameters:

• input_file (required): Path to protein FASTA file
• databases (optional): Comma-separated list of databases to search
• output_format (optional): Output format (tsv, xml, json, gff3) - default: tsv
• goterms (optional): Include GO term annotations - default: true
• pathways (optional): Include pathway annotations - default: true
• priority (optional): Job priority (1-10) - default: 5
• notification_email (optional): Email for job completion notification

Job Management Tools

• get_job_status: Check the status of a background job
• get_job_result: Retrieve results from a completed job
• list_my_jobs: List recent jobs
• cancel_job: Cancel a running job

Example:

Run InterProScan analysis on proteins.fasta with GO terms and pathway mapping

Note: InterProScan jobs run in the background due to their computational intensity. Most jobs take 30 minutes to several hours depending on the input size.

Development

Running tests

pytest tests/

Building Docker image

docker build -t bio-mcp-interpro .

License

MIT License