🧬 BioVis-MCP: Automated Bioinformatics Visualization

"From raw data to publication-ready figures in seconds."

BioVis-MCP is a high-performance Model Context Protocol (MCP) server that empowers Large Language Models (like Claude) with the ability to generate publication-quality (300 DPI) bioinformatics visualizations directly from raw biological data.

No more manual Matplotlib tweaking. Just send the data, and get a verified, manuscript-ready image path.

✨ Features (Phases 1-3)

📊 Visualization Suite

• Volcano Plots: High-resolution visualization of differential expression, with automated significance highlighting (Up/Down regulated).

• PCA Plots: Principal Component Analysis for sample relationship and variance insights.

• Expression Heatmaps: Professional heatmaps with hierarchical clustering and customizable colormaps.

• MA Plots: Classic M-versus-A plots for global genomic trends.

• Pathway Enrichment: Horizontal bar charts and dynamic Bubble Charts (Gene Count vs. Significance).

📝 Reporting & AI Integration

• Smart Figure Captions: Context-aware, statistically accurate scientific captions generated automatically.

• Comprehensive Reports: Multi-figure assembly into professional PDF and DOCX documents.

• High DPI Standards: All figures are generated at 300 DPI using bbox_inches='tight' for Q1 journal compliance.

Related MCP server: GenomeMCP

🛠️ Tech Stack

• Framework: FastMCP

• Libraries: Pandas, Scikit-learn, Scipy, Matplotlib, Seaborn

• Export Formats: PNG (Figures), PDF & DOCX (Reports)

🚀 Installation & Claude Integration

BioVis-MCP can be added to Claude Desktop using one of the following methods.

Method 1: Using uvx (Recommended)

This is the fastest way to run BioVis-MCP without manual installation. Ensure you have uv installed.

Add this to your claude_desktop_config.json:

{
  "mcpServers": {
    "BioVis-MCP": {
      "command": "uvx",
      "args": ["biovis-mcp"]
    }
  }
}

Method 2: Using pip

If you prefer a standard installation:

pip install biovis-mcp

Then add this to your claude_desktop_config.json:

{
  "mcpServers": {
    "BioVis-MCP": {
      "command": "python",
      "args": [
        "-m",
        "biovis_mcp.server"
      ]
    }
  }
}

🛠️ Development & Contributing

If you want to contribute or modify the server locally:

1. Clone the Repository

git clone https://github.com/ZaEyAsa/biovis-mcp.git
cd biovis-mcp

2. Install for Development

pip install -e .[dev]

3. Developer Configuration (Claude Desktop)

For local development, point directly to your server.py:

"BioVis-MCP-Dev": {
  "command": "C:/path/to/python.exe",
  "args": [
    "C:/path/to/biovis-mcp/src/biovis_mcp/server.py"
  ],
  "env": {
    "PYTHONPATH": "C:/path/to/biovis-mcp/src"
  }
}

📖 Available Tools

• generate_volcano_plot(data, title, fc_threshold, pval_threshold)

• generate_bar_enrichment(data, title, top_n, color)

• generate_heatmap_plot(data, title, cluster, cmap)

• generate_pca_plot(data, metadata, title, group_col)

• generate_bubble_enrichment(data, title, top_n)

• generate_ma_plot(data, title, pval_threshold)

• get_figure_caption(tool_type, stats)

• create_report(figures_with_captions, format, report_name)

📁 Output Structure

• /figures: High-resolution PNG files.

• /reports: Formatted PDF and DOCX documents.

Developed by ZaEyAsa — Your Advanced Agentic Bio-Visualization Assistant.