STRING Protein-Protein Interaction Query

Query the STRING database to retrieve protein-protein interactions with comprehensive confidence scores.

When to Use

• Find interaction partners for a protein (by UniProt ID)
• Retrieve confidence scores for PPIs (experimental, text mining, database)
• Build protein interaction networks for pathway analysis
• Identify potential protein complexes or functional modules

Workflow

Basic Query

from open_biomed.tools.tool_registry import TOOLS

# Query STRING for interaction partners
tool = TOOLS["ppi_string_request"]
results, _ = tool.run(uniprot_id="P04637")  # TP53

# Access results
for interaction in results:
    print(f"{interaction['partner_gene']}: {interaction['combined_score']}")

Custom Parameters

# High confidence interactions only, limit to 20
results, _ = tool.run(
    uniprot_id="P04637",
    species=9606,           # Human (default)
    required_score=700,     # High confidence (default)
    limit=20                # Max interactors
)

Parameters

Confidence Score Thresholds

Expected Output

[
  {
    "query_protein": "TP53",
    "partner_string_id": "9606.ENSP00000340989",
    "partner_gene": "SFN",
    "combined_score": 0.999,
    "scores": {
      "experimental": 0.981,
      "text_mining": 0.859,
      "database": 0.75,
      "coexpression": 0.0,
      "phylogenetic": 0.0,
      "gene_fusion": 0.0,
      "neighborhood": 0.0
    },
    "ncbi_taxon_id": 9606
  }
]

Score Interpretation

Error Handling

Common Organism IDs

References

• examples/basic_query.py - Complete example script
• references/score_details.md - Detailed score methodology
• STRING API Docs: https://string-db.org/help/api/