get_enrichment_figure
Return URL of enrichment scatter figure
LABSCO SUMMARY
get_version_info matters more than it looks: a STRING figure without its database version is not reproducible.
Every read takes a species and most take required_score, so the same protein set can produce very different networks depending on the confidence cut — that threshold belongs in whatever you write up, not just in the call. get_network_image and get_enrichment_figure return image URLs rather than bytes, with svg, highres and download_image deciding the form, which keeps asking for a picture cheap in context. get_functional_enrichment is the one that accepts background_identifiers, and supplying it is the difference between enrichment against the whole genome and enrichment against the set you actually assayed.
What it is
A STRING database client with 11 tools covering identifier mapping, interaction networks and partners, functional annotation and enrichment, homology, and the rendered network and enrichment figures.
What you get
- map_identifiers resolves protein identifiers to STRING IDs for a species, which is the form the other tools expect.
- get_network_interactions returns interaction edges for a protein set, filtered by network_type and required_score, with add_nodes to widen the network beyond what you named.
- get_interaction_partners returns the partners of given proteins under the same score and network-type filters.
- get_functional_enrichment runs GO and pathway enrichment over a protein set, optionally against a background set you supply through background_identifiers.
- get_functional_annotation returns all functional annotations for proteins, with allow_pubmed and only_pubmed controlling the literature side.
- get_ppi_enrichment returns protein-protein interaction enrichment statistics for a set.
- Homology: get_homology within a species, and get_homology_best for the best matches against a second species.
- Figures as URLs: get_network_image with network_flavor, node colouring, svg and highres, and get_enrichment_figure with a category, colour palette, axis choice and similarity grouping.
- get_version_info returns the STRING database version behind every answer.
Requirements
Nothing — no account, no key.
Setup effort
One command — pip install string-mcp
USE CASES
01
Turning a list of gene symbols into something STRING will accept
Mapping first avoids the silent failure where an unrecognised identifier simply drops out of a network, and the version read pins the answer to a release.
02
Building the interaction network around a protein set
The confidence threshold and the network type decide what an edge means here, and the enrichment statistic says whether the network is denser than chance.
03
Testing whether a hit list is functionally enriched
Supplying the background set is what makes the test about your experiment rather than about the genome, and the figure comes back as a URL.
04
Finding the equivalent protein in another species
Homology within a species and the best match against a second species are separate calls, and both expect mapped identifiers.
05
Getting a figure out for a write-up
Both image tools return a URL rather than bytes, so requesting a picture costs nothing in context — and the version read is what makes it citable.
11 tools
Counted once with only startup setup. Other versions, modes or optional credentials may expose a different list.
get_functional_annotation
Retrieve all functional annotations for proteins.
get_functional_enrichment
Perform GO / pathway enrichment on a protein set.
get_homology
Get homology information for proteins.
get_homology_best
Get best homology matches for proteins.
get_interaction_partners
Retrieve interaction partners for given proteins.
get_network_image
Return URL of STRING network image
get_network_interactions
Retrieve STRING interaction edges.
get_ppi_enrichment
Get protein-protein interaction enrichment statistics.
get_version_info
Return the current STRING database version.
map_identifiers
Map protein identifiers to STRING IDs.
BEFORE YOU INSTALL
What to know first
Tools you get
11 tools when we connected. The documentation's tool list is out of date. get_protein_similarity is not available; get_homology and get_homology_best are, and get_enrichment_figure is undocumented.
Where it runs
The documented client runs the server on your machine.
CONTEXT COST
Light Standing cost
Rated by the size of the 11 tool definitions that ride along on every turn, not by how many there are
INSTALLATION
Connect your client
Server setup
Setup
TERMINAL
git clone https://github.com/MCPmed/STRINGmcp
cd STRINGmcp
pip install -e .pip
TERMINAL
pip install string-mcpMCP Clients
Claude Desktop
CLAUDE DESKTOP · JSON
{
"mcpServers": {
"string-mcp": {
"command": "string-mcp",
"env": {}
}
}
}string-mcp has to be on the client's PATH; otherwise use its full path inside the Python environment you installed into.
