STRING REST API

Base URL

https://string-db.org/api

URL Pattern

/api/{output_format}/{method}

• output_format: json, tsv, tsv-no-header, image, svg (not all formats for all endpoints)
• method: endpoint name (see below)

Authentication

No API key required. All endpoints are public.

Key Endpoints

1. Resolve protein identifiers

Map protein names/identifiers to STRING internal IDs. Always do this first to get canonical STRING IDs.

GET /api/json/get_string_ids?identifiers={query}&species={taxid}

Example:

https://string-db.org/api/json/get_string_ids?identifiers=TP53&species=9606

Response:

[
  {
    "stringId": "9606.ENSP00000269305",
    "preferredName": "TP53",
    "ncbiTaxonId": 9606,
    "taxonName": "Homo sapiens",
    "annotation": "Cellular tumor antigen p53; ..."
  }
]

2. Get interaction partners (network)

GET /api/json/interaction_partners?identifiers={proteins}&species={taxid}

Example:

https://string-db.org/api/json/interaction_partners?identifiers=TP53&species=9606&limit=10&required_score=900

Response:

[
  {
    "stringId_A": "9606.ENSP00000269305",
    "stringId_B": "9606.ENSP00000261842",
    "preferredName_A": "TP53",
    "preferredName_B": "MDM2",
    "ncbiTaxonId": 9606,
    "score": 0.999,
    "nscore": 0,
    "fscore": 0,
    "pscore": 0,
    "ascore": 0.93,
    "escore": 0.994,
    "dscore": 0.9,
    "tscore": 0.981
  }
]

Score channels: nscore (neighborhood), fscore (fusion), pscore (phylogenetic co-occurrence), ascore (co-expression), escore (experimental), dscore (database/curated), tscore (text mining).

3. Get network interactions between a set of proteins

GET /api/json/network?identifiers={proteins}&species={taxid}

Example — network among a set of proteins:

https://string-db.org/api/json/network?identifiers=TP53%0dBRCA1%0dATM%0dCHEK2%0dMDM2&species=9606&required_score=700

Returns all pairwise interactions among the input set.

4. Network image

GET /api/image/network?identifiers={proteins}&species={taxid}
GET /api/svg/network?identifiers={proteins}&species={taxid}

Returns a PNG image or SVG of the interaction network.

Example:

https://string-db.org/api/image/network?identifiers=TP53%0dBRCA1%0dMDM2&species=9606

5. Functional enrichment analysis

Perform Gene Ontology, KEGG pathway, and other enrichment analysis on a set of proteins.

GET /api/json/enrichment?identifiers={proteins}&species={taxid}

Example:

https://string-db.org/api/json/enrichment?identifiers=TP53%0dBRCA1%0dATM%0dCHEK2%0dCDK2%0dCDKN1A&species=9606

Response:

[
  {
    "category": "Process",
    "term": "GO:0006974",
    "description": "cellular response to DNA damage stimulus",
    "number_of_genes": 6,
    "number_of_genes_in_background": 781,
    "ncbiTaxonId": 9606,
    "inputGenes": "TP53,BRCA1,ATM,CHEK2,CDK2,CDKN1A",
    "preferredNames": "TP53,BRCA1,ATM,CHEK2,CDK2,CDKN1A",
    "p_value": 1.2e-12,
    "fdr": 5.6e-10
  }
]

Categories include: Process (GO Biological Process), Function (GO Molecular Function), Component (GO Cellular Component), KEGG, Pfam, InterPro, SMART, Keyword (UniProt), Reactome, WikiPathways, HPO (Human Phenotype Ontology).

6. Get protein annotations/info

GET /api/json/get_string_ids?identifiers={proteins}&species={taxid}

Maps arbitrary names to STRING IDs with annotation text.

Example:

https://string-db.org/api/json/get_string_ids?identifiers=CDK2%0dp53&species=9606

Response:

[
  {
    "queryIndex": 0,
    "queryItem": "CDK2",
    "stringId": "9606.ENSP00000266970",
    "ncbiTaxonId": 9606,
    "taxonName": "Homo sapiens",
    "preferredName": "CDK2",
    "annotation": "Cyclin-dependent kinase 2; ..."
  }
]

7. Get homology / best-hit in another species

GET /api/json/homology_best?identifiers={proteins}&species={taxid}&species_b={taxid_b}

Example:

https://string-db.org/api/json/homology_best?identifiers=TP53&species=9606&species_b=10090

8. PPI enrichment (is my set more connected than expected?)

GET /api/json/ppi_enrichment?identifiers={proteins}&species={taxid}

Example:

https://string-db.org/api/json/ppi_enrichment?identifiers=TP53%0dBRCA1%0dATM%0dCHEK2&species=9606

Response:

[
  {
    "number_of_nodes": 4,
    "number_of_edges": 6,
    "average_node_degree": 3.0,
    "local_clustering_coefficient": 1.0,
    "expected_number_of_edges": 1,
    "p_value": 0.000123
  }
]

Common Species Taxonomy IDs

Rate Limits

• No published hard rate limit, but the API is intended for programmatic access at moderate rates.
• Recommended: max 1 request per second.
• For large-scale data downloads, use the flat-file downloads on the STRING website instead.
• If you send too many requests, you may receive HTTP 429 or temporary blocking.
• Multiple identifiers per request is strongly preferred over multiple single-identifier requests.

Error Handling

• Returns HTTP 400 for malformed requests.
• Returns HTTP 404 if no matching protein is found.
• Empty JSON array [] if the query is valid but returns no results (e.g., no interactions above the threshold).
• Include species parameter whenever possible to avoid ambiguous identifier resolution.

Current STRING API contract also supports regulatory networks. Pin the returned STRING version and network type; best sequence similarity does not by itself establish orthology.