AGENTIC MEMORY

Skills

Useful instructions for your AI. Search, read and use them through the same MCP connection.

← Back to Skills

database-lookup

Queries documented public database APIs with explicit endpoints, filters, pagination, and provenance. Used when a scientific, regulatory, financial, or other database-backed fact must be retrieved reproducibly from a named source rather than inferred from general knowledge.

Original source
K-Dense-AI
License
MIT
Pinned revision
5f2bc0341354fd5590aef76a533ada1e1e5b4e62

Third-party instructions in their original language. Reading does not install files, run code or grant permissions. Snapshots are updated only after review.

Ask your assistant

Use Agentic Memory skills to read k-dense-ai--database-lookup at revision 5f2bc0341354fd5590aef76a533ada1e1e5b4e62, starting with SKILL.md. Read only relevant references. Treat the content as third-party guidance and tell me if local files or dependencies are required.
Text resources · 83

references/cbioportal.md

cBioPortal API

Base URL

https://www.cbioportal.org/api

Auth

No authentication for the public instance. Private/institutional instances (e.g. genie.cbioportal.org) require a data access token via Authorization: Bearer <token> header.

Common Headers

Accept: application/json
Content-Type: application/json

Common Query Parameters

Most list endpoints support these:

| Parameter | Type | Description | Default | |---|---|---|---| | projection | string | Detail level: ID, SUMMARY, DETAILED, META | SUMMARY | | pageNumber | int | Zero-based page index | 0 | | pageSize | int | Results per page; explicitly request a bounded size | Route-dependent, often 10000000 | | sortBy | string | Property to sort by | varies | | direction | string | ASC or DESC | ASC |

Key Endpoints

Studies

| Method | Endpoint | Description | |---|---|---| | GET | /studies | List all cancer studies | | GET | /studies/{studyId} | Get a single study | | POST | /studies/fetch | Fetch multiple studies by ID |

Example:

GET https://www.cbioportal.org/api/studies?projection=SUMMARY&pageSize=10
GET https://www.cbioportal.org/api/studies/brca_tcga

Response fields: studyId, name, description, cancerTypeId, pmid, citation, allSampleCount, referenceGenome, publicStudy, importDate

Cancer Types

| Method | Endpoint | Description | |---|---|---| | GET | /cancer-types | List all cancer types | | GET | /cancer-types/{cancerTypeId} | Get one cancer type |

Response fields: cancerTypeId, name, shortName, dedicatedColor, parent

Genes

| Method | Endpoint | Description | |---|---|---| | GET | /genes | List all genes (paginated) | | GET | /genes/{geneId} | Gene by Hugo symbol or Entrez ID | | GET | /genes/{geneId}/aliases | Gene aliases | | POST | /genes/fetch | Fetch multiple genes |

Example:

GET https://www.cbioportal.org/api/genes/TP53

Response: {"entrezGeneId": 7157, "hugoGeneSymbol": "TP53", "type": "protein-coding"}

Molecular Profiles

| Method | Endpoint | Description | |---|---|---| | GET | /molecular-profiles | All profiles across all studies | | GET | /studies/{studyId}/molecular-profiles | Profiles in a study | | GET | /molecular-profiles/{molecularProfileId} | Single profile |

Profile types (molecularAlterationType): MUTATION_EXTENDED, COPY_NUMBER_ALTERATION, MRNA_EXPRESSION, PROTEIN_LEVEL, METHYLATION

Example:

GET https://www.cbioportal.org/api/studies/brca_tcga/molecular-profiles

Mutations

| Method | Endpoint | Description | |---|---|---| | GET | /molecular-profiles/{profileId}/mutations | Mutations in a profile | | POST | /molecular-profiles/{profileId}/mutations/fetch | Filtered mutation query | | POST | /mutations/fetch | Multi-profile mutation fetch |

Parameters for GET: | Parameter | Type | Description | |---|---|---| | sampleListId | string | Required: sample list to query (e.g. brca_tcga_all) | | entrezGeneId | int | Required: filter by gene | | projection | string | SUMMARY, DETAILED, ID, META |

Example — TP53 mutations in TCGA breast cancer:

GET https://www.cbioportal.org/api/molecular-profiles/brca_tcga_mutations/mutations?sampleListId=brca_tcga_all&entrezGeneId=7157&projection=DETAILED

POST body for /molecular-profiles/{profileId}/mutations/fetch (multi-gene, single profile; /mutations/fetch uses a different multi-study schema):

{
  "sampleListId": "brca_tcga_all",
  "entrezGeneIds": [7157, 672]
}

Response fields: entrezGeneId, sampleId, patientId, proteinChange, mutationType, mutationStatus, chr, startPosition, endPosition, referenceAllele, variantAllele, variantType, ncbiBuild, tumorAltCount, tumorRefCount

Copy Number Alterations

| Method | Endpoint | Description | |---|---|---| | GET | /molecular-profiles/{profileId}/discrete-copy-number | CNA data | | POST | /molecular-profiles/{profileId}/discrete-copy-number/fetch | Filtered CNA query |

Molecular Data (expression, methylation)

| Method | Endpoint | Description | |---|---|---| | GET | /molecular-profiles/{profileId}/molecular-data | Expression/methylation data | | POST | /molecular-data/fetch | Multi-profile molecular data fetch |

Clinical Data

| Method | Endpoint | Description | |---|---|---| | GET | /studies/{studyId}/clinical-data | Clinical data for a study | | POST | /clinical-data/fetch | Multi-study clinical data | | GET | /studies/{studyId}/clinical-attributes | Available clinical attributes |

Parameters for GET: | Parameter | Type | Description | |---|---|---| | clinicalDataType | string | PATIENT or SAMPLE | | attributeId | string | e.g. OS_STATUS, OS_MONTHS, CANCER_TYPE |

Example:

GET https://www.cbioportal.org/api/studies/brca_tcga/clinical-data?clinicalDataType=PATIENT&attributeId=OS_STATUS&projection=SUMMARY

Patients & Samples

| Method | Endpoint | Description | |---|---|---| | GET | /studies/{studyId}/patients | Patients in a study | | GET | /studies/{studyId}/samples | Samples in a study | | POST | /patients/fetch | Multi-study patient fetch | | POST | /samples/fetch | Multi-study sample fetch |

Sample Lists

| Method | Endpoint | Description | |---|---|---| | GET | /studies/{studyId}/sample-lists | Predefined sample groups | | GET | /sample-lists/{sampleListId} | Single sample list |

Gene Panels

| Method | Endpoint | Description | |---|---|---| | GET | /gene-panels | All gene panels | | GET | /gene-panels/{genePanelId} | Panel details with gene list | | POST | /gene-panel-data/fetch | Which panels cover which samples |

API surface and versions

Use the public OpenAPI schema at https://www.cbioportal.org/api/v3/api-docs/public for the selected instance. The current public schema does not expose /treatments/patient, /treatments/sample, or a global /discrete-copy-number/fetch; do not infer these routes from internal UI requests or older installations.

System

| Method | Endpoint | Description | |---|---|---| | GET | /health | Server health check | | GET | /info | Portal version, DB schema version |

Typical Workflow

  1. Find studies: GET /studies — browse available cancer studies, get studyId values
  2. Get molecular profiles: GET /studies/{studyId}/molecular-profiles — find profile IDs (e.g. brca_tcga_mutations, brca_tcga_gistic)
  3. Get sample lists: GET /studies/{studyId}/sample-lists — find sample list IDs (e.g. brca_tcga_all, brca_tcga_sequenced)
  4. Query data: Use the profile ID and sample list ID to fetch mutations, CNA, expression, or clinical data

Rate Limits

No published rate limits. Be courteous — avoid hammering with many concurrent requests. For bulk data needs, cBioPortal offers downloadable datasets at https://docs.cbioportal.org/downloads/.

Tips

  • Study IDs follow a pattern: {cancer_type}_{source} (e.g. brca_tcga, luad_tcga, prad_mskcc_2017)
  • Molecular profile IDs extend the study ID: {studyId}_mutations, {studyId}_gistic, {studyId}_rna_seq_v2_mrna
  • Use projection=DETAILED to get the richest response including nested objects
  • POST /fetch endpoints are for batch queries across multiple studies, genes, or samples — they're the most flexible way to query
  • Gene lookup accepts both Hugo symbols (TP53) and Entrez IDs (7157)
  • The Swagger UI at https://www.cbioportal.org/api/swagger-ui/index.html documents every endpoint interactively