pairix-mcp

An MCP server for pairix — a fast indexer and query tool for bgzipped .pairs files used in Hi-C data analysis.

Tools

Tool	Description
check_pairix_installed	Check whether pairix and bgzip are installed and accessible
install_pairix	Install pairix into the MCP server's Python environment (macOS/Linux)
install_pairix_windows	Install pairix inside WSL on Windows
pairix_index	Build a .px2 index for a bgzipped .pairs file
pairix_query	Query an indexed .pairs file by 1D or 2D genomic region
check_job_status	Poll the status of a background job
list_jobs	List all background jobs in the current session

Typical Hi-C workflow

1. pairix_index(input_pairs_gz="sample.pairs.gz", preset="pairs")
   → sample.pairs.gz.px2

2. pairix_query(
       input_pairs_gz="sample.pairs.gz",
       regions=["chr1:1000000-2000000|chr2:5000000-6000000"]
   )
   → returns matching pair records

Note: The input must already be bgzipped. Use a separate bgzip tool or bgzip CLI before indexing.

Region formats

• 2D region: "chr1:1000000-2000000|chr2:5000000-6000000" — pairs where end1 is in chr1 window AND end2 is in chr2 window
• 1D region: "chr1:1000000-2000000" — pairs where either end falls in the window

Installation

cd pairix-mcp
uv sync
uv run main.py

MCP config

{
  "mcpServers": {
    "pairix-mcp": {
      "command": "uv",
      "args": ["--directory", "/path/to/pairix-mcp", "run", "main.py"]
    }
  }
}