string-database

Query the STRING database for protein interactions, enrichment, and homology.

2.7k|283|Updated May 13, 2026
One-click install
npx skills add https://github.com/google-deepmind/science-skills --skill string-database-google-deepmind
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: string-database
Source: https://github.com/google-deepmind/science-skills/tree/main/skills/string_database
Command: npx skills add https://github.com/google-deepmind/science-skills --skill string-database-google-deepmind

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) components.

What problem does it solve?

This Skill helps you investigate protein-protein interactions and biological meaning by querying the STRING database for interaction partners, network structure, homology, and enrichment results tied to a specific organism.

Core Features & Use Cases

  • Protein interaction queries: Retrieve STRING interaction networks (physical or functional), interaction partners, and confidence-related scores for specified proteins.
  • Functional enrichment: Run GO/KEGG/Pfam enrichment and PPI enrichment to identify significantly overrepresented functional terms in a protein set.
  • Homology and identifier mapping: Map common protein/gene names to STRING IDs for faster queries, then compute homology similarity and best hits across species.
  • Use case example: You have a list of candidate proteins from an experiment and want to know which pathways and functional terms are enriched, while also examining whether key proteins interact directly in a specific species.

Quick Start

Ask the AI: "For species with NCBI Taxon ID 9606, query STRING interactions between TP53 and MDM2 and return the top interaction partners with confidence."

Frequently Asked Questions about string-database

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I find protein-protein interactions for a specific species using STRING?

To find protein-protein interactions, query the STRING database by providing a protein set and the required NCBI Taxon ID. This retrieves interaction partners, network structures, and confidence scores for that specific species.

Can I run GO and KEGG functional enrichment on a list of candidate proteins?

Yes, you can run functional enrichment on candidate proteins. The Skill queries the STRING database to calculate GO, KEGG, and Pfam enrichment, identifying significantly overrepresented functional terms and pathways.

Do I need an NCBI Taxon ID to map protein identifiers to STRING IDs?

Yes, an NCBI Taxon ID is required to map common protein or gene names to STRING IDs. The Skill enforces required NCBI Taxon IDs without guessing species to ensure accurate identifier mapping and homology similarity computation.

What's the best way to retrieve STRING interaction networks without loading large outputs into context?

The best way to retrieve STRING interaction networks is by running bundled scripts via uv run with the --output flag to write results to a file. This reference-guided CLI approach avoids loading large stdout into context.

Does STRING database homology search support finding best hits across different species?

Yes, STRING database homology search supports finding best hits across species. After mapping identifiers to STRING IDs, you can compute homology similarity and retrieve best hits for your specified protein set.

What are the limitations of using STRING for neighborhood network analysis?

A limitation of STRING neighborhood network analysis is that species must be explicitly specified using an NCBI Taxon ID. The Skill will not guess species, so queries lacking this required identifier will fail to return interaction evidence.