playbooks

home / mcp / human protein atlas mcp server

Human Protein Atlas MCP Server

Provides programmatic access to Human Protein Atlas data including expression, localization, and pathology data via MCP.

Installation
Add the following to your MCP client configuration file.

Configuration

View docs
{
  "mcpServers": {
    "augmented-nature-proteinatlas-mcp-server": {
      "command": "node",
      "args": [
        "/path/to/proteinatlas-server/build/index.js"
      ]
    }
  }
}

You can access Human Protein Atlas data through a dedicated MCP (Model Context Protocol) server that exposes protein information, tissue and brain expressions, subcellular localization, pathology data, and antibody details. This server lets you search, retrieve, and compare protein data programmatically, enabling fast, batch-friendly queries for research and education.

How to use

You will run the MCP server locally or securely in your environment and connect your MCP client to it using a stdio configuration. You can also plan to integrate the server into batch workflows and batch lookup tasks to retrieve multiple proteins efficiently. Use the server to search for proteins, fetch detailed protein information, and access expression and localization data across tissues, blood, brain regions, and pathology data. The available tools expose common operations such as searching, retrieving by gene or Ensembl ID, and performing advanced searches across multiple filters.

How to install

Prerequisites You need Node.js 18 or higher and npm or yarn as your package manager.

Setup steps follow a straightforward flow: clone or download the server code, install dependencies, build the project, and start the server.

Concrete steps you can perform now:

# Step 1: Install dependencies
npm install

# Step 2: Build the project
npm run build

# Step 3: Start the server
npm start
# or
node build/index.js

Example client configuration

To integrate this MCP server with your MCP client, use a stdio configuration that runs the server locally.

{
  "mcpServers": {
    "proteinatlas": {
      "command": "node",
      "args": ["/path/to/proteinatlas-server/build/index.js"]
    }
  }
}

Available tools

search_proteins

Search Human Protein Atlas for proteins by name, gene symbol, or description and return results with configurable output formats and columns.

get_protein_info

Fetch detailed information for a specific protein by gene symbol, returning comprehensive data fields.

get_protein_by_ensembl

Retrieve protein information using an Ensembl gene ID to access data tied to the gene's Ensembl entry.

get_tissue_expression

Obtain tissue-specific expression data for a given gene, enabling tissue-focused analyses.

search_by_tissue

Find proteins highly expressed in a specified tissue with optional expression level filters and result limits.

get_blood_expression

Access blood cell expression data for a given protein to study hematologic relevance.

get_brain_expression

Fetch brain region expression data for a protein to explore neural expression patterns.

get_subcellular_location

Retrieve subcellular localization data for a protein, including localization patterns and reliability.

search_by_subcellular_location

Search proteins localized to specific subcellular compartments with optional reliability filters.

get_pathology_data

Obtain cancer and pathology data for a protein, including prognostic and therapeutic relevance.

search_cancer_markers

Identify proteins associated with cancers or with prognostic value based on provided criteria.

advanced_search

Perform an advanced search using multiple filters such as tissue, location, cancer prognostic, and more.

batch_protein_lookup

Look up multiple proteins in a single request with optional column selections.

compare_expression_profiles

Compare expression profiles across several proteins for tissue, brain, blood, or single-cell data.