Scanpy-MCP

What is Scanpy-MCP?

Scanpy-MCP is a natural language interface for single-cell RNA-Seq (scRNA-Seq) analysis that wraps Scanpy functions through the Model Context Protocol (MCP). It allows users and agent developers to perform IO, preprocessing, clustering, differential expression, and plotting using plain language.

How to use Scanpy-MCP?

Install via pip install scanpy-mcp and run scanpy-mcp run to start the server. Configure it in any MCP-compatible client (e.g., Cherry Studio, Cline) using either a local command or a remote SHTTP endpoint.

Key features of Scanpy-MCP

• IO module: read and write scRNA-Seq data
• Preprocessing: filtering, QC, normalization, scaling, HVG, PCA, Neighbors
• Tool module: clustering, differential expression
• Plotting module: violin, heatmap, dotplot
• Works with any MCP-supporting AI client or framework
• Supports local and remote (SHTTP) transport

Use cases of Scanpy-MCP

• Perform complete scRNA-Seq analysis using natural language commands
• Integrate Scanpy functions into AI agent applications
• Run analysis interactively in MCP-aware chatbots like Cherry Studio
• Automate single-cell data processing pipelines through agent frameworks

FAQ from Scanpy-MCP

How do I install Scanpy-MCP?

Install from PyPI using pip install scanpy-mcp and run scanpy-mcp run to test.

What transport options does Scanpy-MCP support?

It supports standard input/output (stdio) for local use and SHTTP for remote access.

Which MCP clients can I use with Scanpy-MCP?

It works with any MCP client, such as Cherry Studio, Cline, and agent frameworks like Agno.

What analysis modules are available in Scanpy-MCP?

It provides IO, preprocessing, clustering/differential expression (tools), and plotting modules.