search_proteinsLABSCO SUMMARY
The expression question, answered where it gets asked.
Most visits to the Atlas end in the same three steps: confirm it is the right protein, see where it is expressed, and note its Ensembl identity. Its search, expression and Ensembl lookup tools cover that loop without the site's navigation in between.
What it is
An unofficial Node MCP server over proteinatlas.org, covering protein search and detail, tissue, blood and brain expression, subcellular location, pathology and cancer markers, with lookup by name, gene symbol or Ensembl ID and batch lookup of several proteins at once.
What you get
- Proteins searched by name, gene symbol or description, with a cap on how many results come back
- Detail for a protein, including its classification and cross-references
- Tissue-specific expression profiles
- Lookup by Ensembl ID, for when you are working from a gene list rather than a name
Requirements
Node.js 18 or higher, npm install and npm run build, then run the built server from your client config. No account or key — the data is the Human Protein Atlas's public repository. Unofficial: not affiliated with the Human Protein Atlas.
Setup effort
Build from source — git clone https://github.com/Augmented-Nature/ProteinAtlas-MCP-Server
USE CASES
Identifying the right protein from a partial name
Search by name, gene symbol or description and limit how many candidates come back.
Checking which tissues express it
Pull the tissue-specific expression profile rather than reading it off a figure.
Reading a protein's basics before a meeting
Get its description, classification and cross-references in one call.
Working from a gene list of Ensembl IDs
Go straight from the identifier to the protein and its expression, with no name matching in between.
14 tools
Listed in project docs. Your client may show a different list.
get_protein_infoget_protein_by_ensemblget_tissue_expressionsearch_by_tissueget_blood_expressionget_brain_expressionget_subcellular_locationsearch_by_subcellular_locationget_pathology_datasearch_cancer_markersadvanced_searchbatch_protein_lookupcompare_expression_profilesBEFORE YOU INSTALL
What to know first
Tools you get
14 tools named in the project's documentation.
Requires
Node.js 18 or later
Where it runs
The documented client runs the server on your machine.
INSTALLATION
Connect your client
Server setup
git clone https://github.com/Augmented-Nature/ProteinAtlas-MCP-Server
cd ProteinAtlas-MCP-Server
npm install
npm run buildMCP Clients
Any MCP client
{
"mcpServers": {
"proteinatlas": {
"command": "node",
"args": [
"<YOUR_PATH_TO_PROTEIN_ATLAS_MCP_SERVER>/build/index.js"
]
}
}
}<YOUR_PATH_TO_PROTEIN_ATLAS_MCP_SERVER> is the absolute path of your ProteinAtlas-MCP-Server clone.