matheus-rech/meta-analysis-mcp-server

Meta-Analysis MCP Server with Cochrane Compliance

A comprehensive Model Context Protocol (MCP) server for conducting systematic reviews and meta-analyses with Cochrane Handbook guidelines and PRISMA 2020 reporting standards.

🎯 Features

Meta-Analysis Tools (7)

• initialize_meta_analysis: Start new meta-analysis project with guided setup
• upload_study_data: Upload and validate study data with comprehensive checks
• perform_meta_analysis: Execute meta-analysis with automated statistical methods
• generate_forest_plot: Create publication-ready forest plots
• assess_publication_bias: Perform publication bias assessment with multiple tests
• generate_report: Create comprehensive meta-analysis reports
• get_session_status: Get current session status and generated files

Cochrane Compliance Tools (4)

• assess_risk_of_bias: Cochrane ROB 2.0 risk of bias assessment
• generate_prisma_checklist: PRISMA 2020 compliance checklist with flowchart
• perform_grade_assessment: GRADE evidence quality assessment
• generate_cochrane_report: Comprehensive Cochrane-compliant systematic review report

🚀 Quick Start

Installation

# Install from source
git clone https://github.com/matheus-rech/meta-analysis-mcp-server.git
cd meta-analysis-mcp-server
pip install -e .

Install R Dependencies

Rscript -e "install.packages(c('meta', 'metafor', 'metaSEM', 'nloptr', 'xml2', 'lme4', 'jsonlite', 'DT', 'rmarkdown', 'knitr', 'ggplot2'), repos='https://cran.rstudio.com/')"

Running the Server

# Run in stdio mode (default)
meta-analysis-mcp-server

# Run in HTTP mode
HTTP_MODE=true PORT=8080 meta-analysis-mcp-server

MCP Client Configuration

Add to your MCP client configuration:

{
  "mcpServers": {
    "meta-analysis": {
      "command": "meta-analysis-mcp-server",
      "args": []
    }
  }
}

📊 Usage Examples

Basic Meta-Analysis Workflow

from extended_protocol_handler import ExtendedMCPProtocolHandler
import asyncio
import json

async def basic_workflow():
    handler = ExtendedMCPProtocolHandler()
    
    # Initialize session
    init_response = await handler.handle_request({
        "method": "tools/call",
        "params": {
            "name": "initialize_meta_analysis",
            "arguments": {
                "user_id": "researcher_001",
                "project_name": "My Systematic Review",
                "study_type": "clinical_trial",
                "effect_measure": "OR"
            }
        }
    })
    
    session_data = json.loads(init_response["content"][0]["text"])
    session_id = session_data["session_id"]
    
    # Upload data and perform analysis...

Cochrane Compliance Workflow

# After basic meta-analysis, add Cochrane compliance
rob_response = await handler.handle_request({
    "method": "tools/call",
    "params": {
        "name": "assess_risk_of_bias",
        "arguments": {
            "session_id": session_id,
            "assessment_type": "automated",
            "rob_version": "rob2"
        }
    }
})

prisma_response = await handler.handle_request({
    "method": "tools/call",
    "params": {
        "name": "generate_prisma_checklist",
        "arguments": {
            "session_id": session_id,
            "review_type": "intervention",
            "include_flowchart": True
        }
    }
})

🏥 Cochrane Handbook Compliance

This server implements guidelines from:

• Cochrane Handbook Chapter 8: Risk of bias assessment using ROB 2.0
• Cochrane Handbook Chapter 10: Statistical analysis and meta-analysis methods
• Cochrane Handbook Chapter 12: Synthesis and presentation of results

📋 PRISMA 2020 Compliance

Includes full PRISMA 2020 checklist with:

• 27-item checklist for systematic reviews
• Flow diagram generation
• Compliance scoring and recommendations
• Integration with systematic review reporting

🔬 GRADE Evidence Assessment

Implements GRADE methodology for:

• Risk of bias evaluation
• Inconsistency assessment
• Indirectness evaluation
• Imprecision analysis
• Publication bias detection

📁 File Structure

meta_analysis_mcp_server/
├── server.py                       # Core MCP server implementation
├── http_server.py                  # HTTP wrapper for MCP server
├── __main__.py                     # Entry point (stdio or HTTP mode)
├── tools/
│   ├── meta_analysis.py            # Meta-analysis tools implementation
│   └── cochrane_compliance.py      # Cochrane compliance tools
├── r_scripts/
│   └── meta_analysis.R             # R statistical analysis scripts
├── templates/
│   ├── report_template.Rmd         # Original report template
│   └── cochrane_report_template.Rmd # Cochrane-compliant template

🌐 Deployment

Docker

docker build -t meta-analysis-mcp-server .
docker run -p 8080:8080 meta-analysis-mcp-server

Railway

Deploy directly from GitHub repository using the included railway.json configuration.

Fly.io

fly deploy

🧪 Testing

Test Original Functionality

python3 demo_workflow.py

Test Cochrane Compliance Features

python3 demo_cochrane_workflow.py

Test Individual Tools

python3 test_individual_tools.py

Development

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

# Run tests
pytest

# Format code
black .

# Type checking
mypy .

🤝 Contributing

This server maintains backward compatibility with all original functionality while adding Cochrane compliance features. When contributing:

1. Preserve all existing tool functionality
2. Follow Cochrane Handbook guidelines
3. Maintain PRISMA 2020 compliance
4. Include comprehensive testing

📚 References

1. Higgins JPT, Thomas J, Chandler J, et al. Cochrane Handbook for Systematic Reviews of Interventions version 6.3. Cochrane, 2022.
2. Page MJ, McKenzie JE, Bossuyt PM, et al. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. BMJ 2021;372:n71.
3. Guyatt GH, Oxman AD, Vist GE, et al. GRADE: an emerging consensus on rating quality of evidence and strength of recommendations. BMJ 2008;336:924-6.

📄 License

MIT License - see LICENSE file for details.

Meta-Analysis MCP Server with Cochrane Compliance - Democratizing systematic reviews through guided workflows and automated R execution.