Skip to main content
Glama

STRING-MCP

A comprehensive Python package for interacting with the STRING database API through a Model Context Protocol (MCP) bridge.

Installation

Install the package in development mode:

pip install -e .

Or install from PyPI (when available):

pip install string-mcp

Related MCP server: STRING-db MCP Server

Claude config

  "mcpServers": {
    "string-mcp": {
      "command": "/path/to/python/env/bin/string-mcp-server",
      "env": {}
    }
  }
}

Usage

MCP Server (Primary Use Case)

The package provides an MCP server for integration with MCP-compatible clients:

# Run the MCP server
string-mcp-server

The MCP server provides the following tools:

  • map_identifiers: Map protein identifiers to STRING IDs

  • get_network_interactions: Get network interactions data

  • get_functional_enrichment: Perform functional enrichment analysis

  • get_network_image: Generate network visualization images

  • get_version_info: Get STRING database version information

Command Line Interface

The package also provides a string-mcp command for standalone usage:

# Run demo
string-mcp demo

# Get help
string-mcp --help

# Map protein identifiers
string-mcp map TP53 BRCA1 EGFR --species 9606

# Get network interactions
string-mcp network TP53 BRCA1 --species 9606

# Generate network image
string-mcp image TP53 BRCA1 --output network.png --species 9606

Python API

from stringmcp.main import StringDBBridge

# Initialize the bridge
bridge = StringDBBridge()

# Map protein identifiers
proteins = ["TP53", "BRCA1", "EGFR"]
mapped = bridge.map_identifiers(proteins, species=9606)  # 9606 = human

# Get network interactions
interactions = bridge.get_network_interactions(proteins, species=9606)

# Perform functional enrichment
enrichment = bridge.get_functional_enrichment(proteins, species=9606)

Features

  • Protein Identifier Mapping: Convert various protein identifiers to STRING IDs

  • Network Analysis: Retrieve protein-protein interaction networks

  • Functional Enrichment: Perform gene ontology and pathway enrichment analysis

  • Network Visualization: Generate network images in various formats

  • Interaction Partners: Find all interaction partners for proteins

  • Functional Annotations: Get detailed functional annotations

  • Protein Similarity: Calculate similarity scores between proteins

  • PPI Enrichment: Test for protein-protein interaction enrichment

  • MCP Integration: Full Model Context Protocol server implementation

API Methods

Core Methods

  • map_identifiers(): Map protein identifiers to STRING IDs

  • get_network_interactions(): Get network interaction data

  • get_network_image(): Generate network visualization images

  • get_interaction_partners(): Find all interaction partners

  • get_functional_enrichment(): Perform enrichment analysis

  • get_functional_annotation(): Get functional annotations

  • get_protein_similarity(): Calculate similarity scores

  • get_ppi_enrichment(): Test for PPI enrichment

  • get_version_info(): Get STRING database version

Configuration

The package uses a StringConfig class for configuration:

from stringmcp.main import StringConfig, StringDBBridge

config = StringConfig(
    base_url="https://string-db.org/api",
    version_url="https://version-12-0.string-db.org/api",
    caller_identity="my_app",
    request_delay=1.0  # Delay between requests in seconds
)

bridge = StringDBBridge(config)

Output Formats

The package supports multiple output formats:

  • JSON: Structured data (default)

  • TSV: Tab-separated values

  • XML: XML format

  • IMAGE: Network visualization images

  • SVG: Scalable vector graphics

  • PSI_MI: PSI-MI format

Species Support

The package supports all species available in STRING. Common species IDs:

  • Human: 9606

  • Mouse: 10090

  • Rat: 10116

  • Yeast: 4932

  • E. coli: 511145

MCP Server Configuration

To use the MCP server with an MCP client, configure it as follows:

{
  "mcpServers": {
    "string-mcp": {
      "command": "string-mcp-server",
      "env": {}
    }
  }
}

The server will automatically handle:

  • JSON-RPC communication

  • Tool discovery and invocation

  • Error handling and reporting

  • Base64 encoding for image data

Development

Setup Development Environment

# Install in development mode with dev dependencies
pip install -e .[dev]

# Format code
black stringmcp/

# Type checking
mypy stringmcp/

# Lint code
flake8 stringmcp/

Note: Test files are not currently included in this repository. To add tests, create a tests/ directory and add test files following the pytest configuration in pyproject.toml.

Project Structure

STRINGmcp/
├── pyproject.toml          # Package configuration and dependencies
├── README.md              # This file
├── LICENSE                # MIT License
├── .gitignore             # Git ignore patterns
├── stringmcp/             # Main package
│   ├── __init__.py        # Package initialization
│   └── main.py            # Core STRING API bridge and MCP server
└── string_mcp.egg-info/   # Package metadata (generated during install)
    ├── PKG-INFO           # Package information
    ├── SOURCES.txt        # Source files list
    ├── dependency_links.txt
    ├── entry_points.txt   # CLI entry points
    ├── requires.txt       # Dependencies
    └── top_level.txt      # Top-level package names

License

MIT License - see LICENSE file for details.

Contributing

  1. Fork the repository

  2. Create a feature branch

  3. Make your changes

  4. Add tests

  5. Run the test suite

  6. Submit a pull request

Support

For issues and questions, please use the GitHub issue tracker.

Example Usage

Complete DNA Repair Protein Analysis

This example demonstrates the comprehensive functionality of the STRING-DB MCP bridge by analyzing a set of well-known human DNA repair proteins: TP53, BRCA1, BRCA2, ATM, and ATR.

2. Protein Identifier Mapping

Map gene symbols to STRING identifiers:

[
  {
    "queryIndex": 0,
    "queryItem": "TP53",
    "stringId": "9606.ENSP00000269305",
    "ncbiTaxonId": 9606,
    "taxonName": "Homo sapiens",
    "preferredName": "TP53",
    "annotation": "Cellular tumor antigen p53; Acts as a tumor suppressor in many tumor types; induces growth arrest or apoptosis depending on the physiological circumstances and cell type..."
  },
  {
    "queryIndex": 1,
    "queryItem": "BRCA1",
    "stringId": "9606.ENSP00000418960",
    "ncbiTaxonId": 9606,
    "taxonName": "Homo sapiens",
    "preferredName": "BRCA1",
    "annotation": "Breast cancer type 1 susceptibility protein; E3 ubiquitin-protein ligase that specifically mediates the formation of 'Lys-6'-linked polyubiquitin chains..."
  },
  {
    "queryIndex": 2,
    "queryItem": "BRCA2",
    "stringId": "9606.ENSP00000369497",
    "ncbiTaxonId": 9606,
    "taxonName": "Homo sapiens",
    "preferredName": "BRCA2",
    "annotation": "Breast cancer type 2 susceptibility protein; Involved in double-strand break repair and/or homologous recombination..."
  },
  {
    "queryIndex": 3,
    "queryItem": "ATM",
    "stringId": "9606.ENSP00000278616",
    "ncbiTaxonId": 9606,
    "taxonName": "Homo sapiens",
    "preferredName": "ATM",
    "annotation": "Serine-protein kinase ATM; Serine/threonine protein kinase which activates checkpoint signaling upon double strand breaks..."
  },
  {
    "queryIndex": 4,
    "queryItem": "ATR",
    "stringId": "9606.ENSP00000343741",
    "ncbiTaxonId": 9606,
    "taxonName": "Homo sapiens",
    "preferredName": "ATR",
    "annotation": "Serine/threonine-protein kinase ATR; Serine/threonine protein kinase which activates checkpoint signaling upon genotoxic stresses..."
  }
]

3. Protein-Protein Interaction Network

Examine network interactions between these proteins:

[
  {
    "stringId_A": "9606.ENSP00000269305",
    "stringId_B": "9606.ENSP00000369497",
    "preferredName_A": "TP53",
    "preferredName_B": "BRCA2",
    "score": 0.995
  },
  {
    "stringId_A": "9606.ENSP00000269305",
    "stringId_B": "9606.ENSP00000343741",
    "preferredName_A": "TP53",
    "preferredName_B": "ATR",
    "score": 0.996
  },
  {
    "stringId_A": "9606.ENSP00000269305",
    "stringId_B": "9606.ENSP00000278616",
    "preferredName_A": "TP53",
    "preferredName_B": "ATM",
    "score": 0.999
  },
  {
    "stringId_A": "9606.ENSP00000269305",
    "stringId_B": "9606.ENSP00000418960",
    "preferredName_A": "TP53",
    "preferredName_B": "BRCA1",
    "score": 0.999
  },
  {
    "stringId_A": "9606.ENSP00000278616",
    "stringId_B": "9606.ENSP00000369497",
    "preferredName_A": "ATM",
    "preferredName_B": "BRCA2",
    "score": 0.995
  },
  {
    "stringId_A": "9606.ENSP00000278616",
    "stringId_B": "9606.ENSP00000418960",
    "preferredName_A": "ATM",
    "preferredName_B": "BRCA1",
    "score": 0.999
  },
  {
    "stringId_A": "9606.ENSP00000278616",
    "stringId_B": "9606.ENSP00000343741",
    "preferredName_A": "ATM",
    "preferredName_B": "ATR",
    "score": 0.999
  },
  {
    "stringId_A": "9606.ENSP00000343741",
    "stringId_B": "9606.ENSP00000369497",
    "preferredName_A": "ATR",
    "preferredName_B": "BRCA2",
    "score": 0.831
  },
  {
    "stringId_A": "9606.ENSP00000343741",
    "stringId_B": "9606.ENSP00000418960",
    "preferredName_A": "ATR",
    "preferredName_B": "BRCA1",
    "score": 0.996
  },
  {
    "stringId_A": "9606.ENSP00000369497",
    "stringId_B": "9606.ENSP00000418960",
    "preferredName_A": "BRCA2",
    "preferredName_B": "BRCA1",
    "score": 0.999
  }
]

Key Findings: All interactions show very high confidence scores (>0.8), with most exceeding 0.99, indicating these proteins form a tightly interconnected functional module.

4. Network Statistics

Check if this network is significantly enriched for interactions:

{
  "number_of_nodes": 5,
  "number_of_edges": 10,
  "average_node_degree": 4.0,
  "local_clustering_coefficient": 1.0,
  "expected_number_of_edges": 5,
  "p_value": 0.0122
}

Statistical Significance: The network shows perfect clustering (coefficient = 1.0) and is significantly enriched for interactions (p = 0.0122), with twice as many edges as expected by chance.

5. Functional Enrichment Analysis

Analyze which biological pathways are enriched in this protein set:

Top DNA Repair Pathways (Selected Results):

[
  {
    "category": "Process",
    "term": "GO:0071479",
    "number_of_genes": 5,
    "preferredNames": ["TP53", "ATM", "ATR", "BRCA2", "BRCA1"],
    "p_value": 9.72e-13,
    "fdr": 1.52e-08,
    "description": "Cellular response to ionizing radiation"
  },
  {
    "category": "Process",
    "term": "GO:0042770",
    "number_of_genes": 5,
    "preferredNames": ["TP53", "ATM", "ATR", "BRCA2", "BRCA1"],
    "p_value": 1.69e-11,
    "fdr": 1.32e-07,
    "description": "Signal transduction in response to DNA damage"
  },
  {
    "category": "Process",
    "term": "GO:0006281",
    "number_of_genes": 5,
    "preferredNames": ["TP53", "ATM", "ATR", "BRCA2", "BRCA1"],
    "p_value": 1.05e-08,
    "fdr": 1.10e-05,
    "description": "DNA repair"
  },
  {
    "category": "KEGG",
    "term": "hsa03440",
    "number_of_genes": 3,
    "preferredNames": ["ATM", "BRCA2", "BRCA1"],
    "p_value": 8.34e-08,
    "fdr": 2.80e-05,
    "description": "Homologous recombination"
  },
  {
    "category": "KEGG",
    "term": "hsa04115",
    "number_of_genes": 3,
    "preferredNames": ["TP53", "ATM", "ATR"],
    "p_value": 5.27e-07,`
    "fdr": 5.44e-05,`
    "description": "p53 signaling pathway"
  }
]

Disease Associations:

[
  {
    "category": "DISEASES",
    "term": "DOID:1612",
    "number_of_genes": 4,
    "preferredNames": ["TP53", "ATM", "BRCA2", "BRCA1"],
    "p_value": 5.72e-10,
    "fdr": 2.02e-06,
    "description": "Breast cancer"
  },
  {
    "category": "DISEASES",
    "term": "DOID:3012",
    "number_of_genes": 3,
    "preferredNames": ["TP53", "BRCA2", "BRCA1"],
    "p_value": 6.59e-10,
    "fdr": 2.02e-06,
    "description": "Li-Fraumeni syndrome"
  }
]

The package can generate protein interaction network visualizations showing evidence-based functional associations.

Example Network Visualization: View Protein Interaction Network DNA Repair Protein Network This visualization shows the protein-protein interaction network for TP53, BRCA1, BRCA2, ATM, and ATR with high-confidence interactions (score ≥ 400).

7. Functional Enrichment Visualization

The package can also create enrichment scatter plots showing the most significantly enriched biological processes.

Example Enrichment Visualization: View Functional Enrichment Plot Functional Enrichment Plot

This visualization displays the top 10 most significantly enriched biological processes and pathways for the DNA repair protein set, showing p-values and gene counts for each enriched term.

Summary

This comprehensive analysis demonstrates that the STRING-DB MCP bridge successfully:

  • Identified all 5 DNA repair proteins with detailed annotations

  • Discovered 10 high-confidence protein interactions (all >0.8 score)

  • Revealed significant pathway enrichments with p-values < 1e-8

  • Confirmed statistical significance of the network (p = 0.0122)

  • Generated both network and enrichment visualizations

The results validate these proteins as a core DNA damage response module, with exceptionally strong enrichment for:

  • Cellular response to ionizing radiation (p = 1.52e-8)

  • DNA damage signaling (p = 1.32e-7)

  • Homologous recombination (p = 2.8e-5)

  • p53 signaling pathway (p = 5.44e-5)

  • Breast cancer associations (p = 2.02e-6)v This showcases the complete functionality of the STRING-DB MCP bridge for protein interaction network analysis and functional annotation.

Available Tools

11 tools
get_enrichment_figureC

Return URL of enrichment scatter figure

Inputs: identifiers (required), species (required), category, group_by_similarity, +6 more.

get_functional_annotationC

Retrieve all functional annotations for proteins.

Inputs: identifiers (required), species, allow_pubmed, only_pubmed.

get_functional_enrichmentC

Perform GO / pathway enrichment on a protein set.

Inputs: identifiers (required), species, background_identifiers.

get_homologyB

Get homology information for proteins.

Inputs: identifiers (required), species.

get_homology_bestC

Get best homology matches for proteins.

Inputs: identifiers (required), species, species_b.

get_interaction_partnersC

Retrieve interaction partners for given proteins.

Inputs: identifiers (required), species, limit, required_score, +1 more.

get_network_imageC

Return URL of STRING network image

Inputs: identifiers (required), species, highres, svg, +6 more.

get_network_interactionsC

Retrieve STRING interaction edges.

Inputs: identifiers (required), species, required_score, add_nodes, +1 more.

get_ppi_enrichmentC

Get protein-protein interaction enrichment statistics.

Inputs: identifiers (required), species, required_score.

get_version_infoB

Return the current STRING database version.

Inputs: no named parameters.

map_identifiersC

Map protein identifiers to STRING IDs.

Inputs: identifiers (required), species, echo_query, limit.

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections.

Loading tool schema changelog…

    TDQS

    B3.3/5.0

    Scored across 11 tools

    Disambiguation4/5

    Most tools have clearly distinct purposes, such as get_functional_annotation for annotations and get_network_interactions for edges. However, get_homology and get_homology_best could be confused, as both retrieve homology information with only subtle differences in scope.

    Naming Consistency5/5

    All tools follow a consistent verb_noun pattern with 'get_' or 'map_' prefixes, such as get_enrichment_figure and map_identifiers. There are no deviations in naming conventions, making the set highly predictable and readable.

    Tool Count5/5

    With 11 tools, the count is well-scoped for a protein interaction and enrichment analysis server. Each tool appears to serve a specific, non-redundant function, fitting within the typical 3-15 tool range for such domains.

    Completeness4/5

    The tool set covers core protein analysis workflows, including enrichment, homology, interactions, and identifier mapping. A minor gap exists in lacking update or delete operations, but this is reasonable for a retrieval-focused database server, and agents can work around this.

    Maintenance

    ActivityInactive
    ResponsivenessOutdated data

    Related MCP Connectors

    Related MCP Servers

    • A
      license
      B
      quality
      D
      maintenance
      A Model Context Protocol server that enhances language models with protein structure analysis capabilities, enabling detailed active site analysis and disease-related protein searches through established protein databases.
      2
      20
      ISC
    • A
      license
      B
      quality
      D
      maintenance
      A comprehensive Model Context Protocol (MCP) server for accessing the STRING protein interaction database. This server provides powerful tools for protein network analysis, functional enrichment, and comparative genomics through the STRING API.
      6
      4
      MIT
    • A
      license
      B
      quality
      D
      maintenance
      Provides access to the STRING protein-protein interaction database for mapping identifiers, retrieving interaction networks, and performing functional enrichment analysis. It enables users to explore protein partners, pathways, and cross-species homology through natural language interactions.
      9
      1
      ISC