🧬 Pathway MCP Agent

A Model Context Protocol (MCP) server for gene enrichment analysis and protein-protein interaction queries.

Features

• 🔬 Gene Enrichment Analysis - Query Enrichr API for pathway enrichment
• 🔗 Protein Interaction - Query STRING-db for gene interactions
• 📊 Visualization - Generate publication-ready bar plots
• 📁 File Support - Read gene lists from CSV, TSV, Excel files
• 🤖 MCP Compatible - Works with Claude Desktop, VS Code, and other MCP clients

Installation

# Clone the repository
git clone https://github.com/YOUR_USERNAME/pathway-mcp-agent.git
cd pathway-mcp-agent

# Install with uv (recommended)
uv sync

# Or install with pip
pip install -e .

Quick Start

Run the MCP Server

uv run pathway-agent

Configure Claude Desktop

Add to %APPDATA%\Claude\claude_desktop_config.json (Windows) or ~/Library/Application Support/Claude/claude_desktop_config.json (macOS):

{
  "mcpServers": {
    "pathway-agent": {
      "command": "uv",
      "args": [
        "--directory",
        "/path/to/pathway-mcp-agent",
        "run",
        "pathway-agent"
      ]
    }
  }
}

Available Tools

perform_enrichment

Perform pathway enrichment analysis on a gene list.

Input: gene_list=["TP53", "BRCA1", "EGFR"], database="KEGG_2021_Human"
Output: Enriched pathways with p-values and gene overlaps

enrichment_with_plot

Perform enrichment analysis and generate a visualization.

Input: gene_list=["TP53", "BRCA1"], database="KEGG_2021_Human", output_path="./plot.png"
Output: Enrichment results + saved bar plot

explain_mechanism

Explain the interaction mechanism between two genes using STRING-db.

Input: gene_a="TP53", gene_b="MDM2"
Output: Interaction score, evidence channels, interpretation

get_gene_partners

Get top interaction partners for a gene.

Input: gene="TP53", limit=10
Output: Top 10 interaction partners with confidence scores

analyze_gene_file

Read genes from a file and perform enrichment analysis.

Input: file_path="./genes.csv", database="KEGG_2021_Human"
Output: File info + enrichment results

list_databases

List all supported enrichment databases.

Supported Databases

Database	Description
KEGG_2021_Human	KEGG Pathways
GO_Biological_Process_2021	Gene Ontology BP
GO_Cellular_Component_2021	Gene Ontology CC
GO_Molecular_Function_2021	Gene Ontology MF
MSigDB_Hallmark_2020	MSigDB Hallmark Gene Sets

Example Output

Enrichment Analysis

Enrichr result for KEGG_2021_Human
Database: KEGG_2021_Human
Gene count: 5
Enriched term count: 110

Top 5 terms:
1. **Breast cancer**
   - P-value: 2.00e-11
   - gene: MYC, KRAS, BRCA1, TP53, EGFR

Gene Interaction

## Interaction: TP53 ↔ MDM2

### Combined Score: 0.999 (highest confidence)

### Evidence Channels:
- Experimental: 0.999
- Database: 0.900
- Text Mining: 0.999

API References

• Enrichr - Ma'ayan Lab
• STRING-db - Protein-Protein Interaction Database