Knowledge Graph

Free

Search, explore, and traverse 855,000+ cancer biology entities and 7.7 million relationships — every one traced to its original data source with a Pearl Causal Hierarchy layer.

Only in Galen

Every relationship is annotated with a Pearl Causal Hierarchy layer (L1 association, L2 intervention, L3 counterfactual) and traced to its original data source. Unlike LLM-generated answers, every fact is provenance-grounded and verifiable. This is not summarized from the internet — it is structured, validated, experimental evidence from peer-reviewed biomedical databases.

GET/entities/{name}

Free

Get Entity Profile

Retrieve the full profile for a cancer biology entity including metadata, relationships, and evidence provenance.

Only in Galen

Parameters

Path Parameters

Name Type Req Description

Name	Type	Req	Description
`name`	string		Entity name — gene name (EGFR), drug name (erlotinib), disease (NSCLC), pathway (MAPK), mutation (BRAF_V600E), etc. Example: `EGFR`

name

string

Entity name — gene name (EGFR), drug name (erlotinib), disease (NSCLC), pathway (MAPK), mutation (BRAF_V600E), etc.

Example: EGFR

Query Parameters

Name Type Req Description

Name	Type	Req	Description
`max_relationships`	integer		Maximum number of relationships to include in the response. Set to 0 for entity metadata only. Default: `50` Example: `100`

max_relationships

integer

Maximum number of relationships to include in the response. Set to 0 for entity metadata only.

Default: 50

Example: 100

Response Schema

idstringUnique entity identifier

namestringHuman-readable entity name

entity_typestringgene, drug, disease, pathway, mutation, protein, cell_line, or clinical_trial

propertiesobjectEntity-specific metadata — molecular weight for drugs, chromosome location for genes, etc.

confidencefloatConfidence score from 0.0 to 1.0 based on evidence quality

provenancestringOriginal data source that contributed this entity

relationship_countintegerTotal relationships across all types

pch_summaryobjectDistribution of relationships across Pearl Causal Hierarchy layers

├──l1_countintegerNumber of L1 (associational) relationships

├──l2_countintegerNumber of L2 (interventional) relationships

└──l3_countintegerNumber of L3 (counterfactual) relationships

relationshipsarrayTyped relationships to other entities

├──source_idstringSource entity identifier

├──target_idstringTarget entity identifier

├──relationship_typestringEdge type (e.g. targets, inhibits, mutated_in)

├──confidencefloatRelationship confidence score (0.0–1.0)

├──pch_layerintegerPearl Causal Hierarchy layer (1, 2, or 3)

├──evidence_typestringType of supporting evidence

└──provenancestringData source for this relationship

Example Request

import requests

resp = requests.get(
    "https://research.usegalen.com/api/v1/entities/EGFR",
    params={"max_relationships": 100},
    headers={"X-API-Key": "YOUR_API_KEY"},
)
entity = resp.json()
print(f"{entity['name']} ({entity['entity_type']}) — {entity['relationship_count']} relationships")
print(f"  PCH: L1={entity['pch_summary']['l1_count']}, L2={entity['pch_summary']['l2_count']}, L3={entity['pch_summary']['l3_count']}")

Example Response

{
  "id": "gene:egfr",
  "name": "EGFR",
  "entity_type": "gene",
  "properties": {
    "full_name": "Epidermal Growth Factor Receptor",
    "chromosome": "7p11.2",
    "uniprot_id": "P00533",
    "gene_family": "Receptor tyrosine kinase",
    "oncokb_annotated": true
  },
  "confidence": 0.97,
  "provenance": "cosmic",
  "relationship_count": 1342,
  "pch_summary": {
    "l1_count": 876,
    "l2_count": 389,
    "l3_count": 77
  },
  "relationships": [
    {
      "source_id": "drug:erlotinib",
      "target_id": "gene:egfr",
      "relationship_type": "targets",
      "confidence": 0.96,
      "pch_layer": 2,
      "evidence_type": "experimental",
      "provenance": "chembl_36"
    },
    {
      "source_id": "gene:egfr",
      "target_id": "disease:nsclc",
      "relationship_type": "driver_in",
      "confidence": 0.94,
      "pch_layer": 2,
      "evidence_type": "clinical_genomics",
      "provenance": "cbioportal"
    },
    {
      "source_id": "gene:egfr",
      "target_id": "pathway:mapk_signaling",
      "relationship_type": "member_of",
      "confidence": 0.99,
      "pch_layer": 1,
      "evidence_type": "curated",
      "provenance": "reactome"
    },
    {
      "source_id": "mutation:egfr_l858r",
      "target_id": "gene:egfr",
      "relationship_type": "variant_in_gene",
      "confidence": 0.98,
      "pch_layer": 1,
      "evidence_type": "curated",
      "provenance": "cosmic"
    },
    {
      "source_id": "drug:osimertinib",
      "target_id": "gene:egfr",
      "relationship_type": "targets",
      "confidence": 0.97,
      "pch_layer": 2,
      "evidence_type": "experimental",
      "provenance": "chembl_36"
    }
  ]
}

Try It

API KeyGet your free API key →

Path Parameters

namestringrequired

Entity name — gene name (EGFR), drug name (erlotinib), disease (NSCLC), pathway (MAPK), mutation (BRAF_V600E), etc.

Query Parameters

max_relationshipsinteger

Maximum number of relationships to include in the response. Set to 0 for entity metadata only.

GET https://research.usegalen.com/api/v1/entities/{name}

Related Endpoints

Search Entities Get Entity Relationships Extract Subgraph

50–500ms

GET/entities/search

Free

Search Entities

Search the knowledge graph for entities matching a query string.

Only in Galen

Searches across 855,000+ structured cancer biology entities from 10+ databases — not web results or LLM memory. Every result is a verified entity with typed relationships and evidence provenance.

Parameters

Query Parameters

Name	Type	Description
`q`	string	Search query — matches entity names by prefix. Case-insensitive. Example: `KRAS`
`entity_type`	string	Filter by entity type genedrugdiseasepathwaymutationproteincell_lineclinical_trial
`limit`	integer	Maximum results to return Default: `20` Example: `10`
`offset`	integer	Number of results to skip for pagination Default: `0`

Response Schema

resultsarrayMatching entities ordered by relevance

├──idstringUnique entity identifier

├──namestringHuman-readable entity name

├──entity_typestringEntity type classification

├──confidencefloatConfidence score (0.0–1.0)

└──relationship_countintegerTotal relationships for this entity

totalintegerTotal matching entities across all pages

querystringThe search query that was executed

Example Request

import requests

resp = requests.get(
    "https://research.usegalen.com/api/v1/entities/search",
    params={"q": "KRAS", "entity_type": "gene", "limit": 10},
    headers={"X-API-Key": "YOUR_API_KEY"},
)
data = resp.json()
for entity in data["results"]:
    print(f"{entity['name']} ({entity['entity_type']}) — {entity['relationship_count']} rels")

Example Response

{
  "results": [
    {
      "id": "gene:kras",
      "name": "KRAS",
      "entity_type": "gene",
      "confidence": 0.98,
      "relationship_count": 2104
    },
    {
      "id": "mutation:kras_g12c",
      "name": "KRAS_G12C",
      "entity_type": "mutation",
      "confidence": 0.96,
      "relationship_count": 487
    },
    {
      "id": "mutation:kras_g12d",
      "name": "KRAS_G12D",
      "entity_type": "mutation",
      "confidence": 0.95,
      "relationship_count": 412
    }
  ],
  "total": 3,
  "query": "KRAS"
}

Try It

API KeyGet your free API key →

Query Parameters

qstringrequired

Search query — matches entity names by prefix. Case-insensitive.

entity_typestring

Filter by entity type

limitinteger

Maximum results to return

offsetinteger

Number of results to skip for pagination

GET https://research.usegalen.com/api/v1/entities/search

Related Endpoints

Get Entity Profile Get Entity Relationships

20–200ms

GET/entities/{name}/relationships

Free

Get Entity Relationships

Retrieve relationships for an entity with filtering by type, PCH layer, and direction.

Only in Galen

Filter relationships by Pearl Causal Hierarchy layer to distinguish statistical associations (L1) from experimental interventions (L2) from counterfactual reasoning (L3). No other API annotates cancer biology relationships with formal causal hierarchy levels.

Parameters

Path Parameters

Name Type Req Description

Name	Type	Req	Description
`name`	string		Entity name to retrieve relationships for Example: `EGFR`

name

string

Entity name to retrieve relationships for

Example: EGFR

Query Parameters

Name	Type	Description
`relationship_type`	string	Filter by relationship type — e.g. targets, inhibits, mutated_in, member_of
`pch_layer`	integer	Filter by Pearl Causal Hierarchy layer 123
`direction`	string	Filter by edge direction relative to the entity Default: `both` bothincomingoutgoing
`limit`	integer	Maximum number of relationships to return Default: `50`
`offset`	integer	Number of results to skip for pagination Default: `0`

Response Schema

relationshipsarrayTyped relationships with evidence metadata

├──idstringUnique relationship identifier

├──source_idstringSource entity identifier

├──source_namestringSource entity display name

├──target_idstringTarget entity identifier

├──target_namestringTarget entity display name

├──relationship_typestringEdge type (e.g. targets, inhibits, mutated_in)

├──confidencefloatRelationship confidence score (0.0–1.0)

├──pch_layerintegerPearl Causal Hierarchy layer (1, 2, or 3)

├──evidence_typestringType of supporting evidence

└──provenancestringData source for this relationship

totalintegerTotal matching relationships across all pages

entity_idstringThe queried entity identifier

Example Request

import requests

resp = requests.get(
    "https://research.usegalen.com/api/v1/entities/EGFR/relationships",
    params={"pch_layer": 2, "direction": "incoming", "limit": 20},
    headers={"X-API-Key": "YOUR_API_KEY"},
)
data = resp.json()
for rel in data["relationships"]:
    print(f"{rel['source_name']} —[{rel['relationship_type']}]→ {rel['target_name']} (L{rel['pch_layer']}, {rel['provenance']})")

Example Response

{
  "relationships": [
    {
      "id": "rel:drug_egfr_erlotinib_targets_001",
      "source_id": "drug:erlotinib",
      "source_name": "erlotinib",
      "target_id": "gene:egfr",
      "target_name": "EGFR",
      "relationship_type": "targets",
      "confidence": 0.96,
      "pch_layer": 2,
      "evidence_type": "experimental",
      "provenance": "chembl_36"
    },
    {
      "id": "rel:drug_egfr_gefitinib_targets_002",
      "source_id": "drug:gefitinib",
      "source_name": "gefitinib",
      "target_id": "gene:egfr",
      "target_name": "EGFR",
      "relationship_type": "targets",
      "confidence": 0.95,
      "pch_layer": 2,
      "evidence_type": "experimental",
      "provenance": "chembl_36"
    },
    {
      "id": "rel:drug_egfr_osimertinib_targets_003",
      "source_id": "drug:osimertinib",
      "source_name": "osimertinib",
      "target_id": "gene:egfr",
      "target_name": "EGFR",
      "relationship_type": "targets",
      "confidence": 0.97,
      "pch_layer": 2,
      "evidence_type": "experimental",
      "provenance": "chembl_36"
    },
    {
      "id": "rel:mut_egfr_l858r_variant_004",
      "source_id": "mutation:egfr_l858r",
      "source_name": "EGFR_L858R",
      "target_id": "gene:egfr",
      "target_name": "EGFR",
      "relationship_type": "variant_in_gene",
      "confidence": 0.98,
      "pch_layer": 1,
      "evidence_type": "curated",
      "provenance": "cosmic"
    },
    {
      "id": "rel:gene_egfr_mapk_member_005",
      "source_id": "gene:egfr",
      "source_name": "EGFR",
      "target_id": "pathway:mapk_signaling",
      "target_name": "MAPK signaling pathway",
      "relationship_type": "member_of",
      "confidence": 0.99,
      "pch_layer": 1,
      "evidence_type": "curated",
      "provenance": "reactome"
    }
  ],
  "total": 1342,
  "entity_id": "gene:egfr"
}

Try It

API KeyGet your free API key →

Path Parameters

namestringrequired

Entity name to retrieve relationships for

Query Parameters

relationship_typestring

Filter by relationship type — e.g. targets, inhibits, mutated_in, member_of

pch_layerinteger

Filter by Pearl Causal Hierarchy layer

directionstring

Filter by edge direction relative to the entity

limitinteger

Maximum number of relationships to return

offsetinteger

Number of results to skip for pagination

GET https://research.usegalen.com/api/v1/entities/{name}/relationships

Related Endpoints

Get Entity Profile Get PCH Annotations Find Paths Between Entities

30–200ms

GET/entities/{name}/subgraph

Free

Extract Subgraph

Extract a multi-hop subgraph centered on an entity from the knowledge graph.

Only in Galen

Extract multi-hop subgraphs from a knowledge graph built on 13 experimental databases. Nodes and edges carry confidence scores and causal hierarchy annotations — giving you a structured, evidence-graded local network, not an LLM-hallucinated diagram.

Parameters

Path Parameters

Name Type Req Description

Name	Type	Req	Description
`name`	string		Entity name to build subgraph around Example: `EGFR`

name

string

Entity name to build subgraph around

Example: EGFR

Query Parameters

Name Type Req Description

Name	Type	Req	Description
`radius`	integer		Number of hops from center entity (1–3) Default: `1` Example: `2`
`max_nodes`	integer		Maximum nodes to include (10–1000) Default: `100`

radius

integer

Number of hops from center entity (1–3)

Default: 1

Example: 2

max_nodes

integer

Maximum nodes to include (10–1000)

Default: 100

Response Schema

centerstringThe center entity identifier

radiusintegerThe traversal radius used

nodesarrayAll entities within the subgraph

├──idstringUnique entity identifier

├──namestringHuman-readable entity name

├──entity_typestringEntity type classification

└──confidencefloatEntity confidence score (0.0–1.0)

edgesarrayAll relationships within the subgraph

├──source_idstringSource entity identifier

├──target_idstringTarget entity identifier

├──relationship_typestringEdge type

├──confidencefloatRelationship confidence score (0.0–1.0)

└──pch_layerintegerPearl Causal Hierarchy layer (1, 2, or 3)

node_countintegerTotal number of nodes in the subgraph

edge_countintegerTotal number of edges in the subgraph

Example Request

import requests

resp = requests.get(
    "https://research.usegalen.com/api/v1/entities/BRAF/subgraph",
    params={"radius": 2, "max_nodes": 100},
    headers={"X-API-Key": "YOUR_API_KEY"},
)
graph = resp.json()
print(f"Subgraph: {graph['node_count']} nodes, {graph['edge_count']} edges (radius={graph['radius']})")
for node in graph["nodes"][:5]:
    print(f"  {node['name']} ({node['entity_type']})")

Example Response

{
  "center": "gene:braf",
  "radius": 2,
  "nodes": [
    {
      "id": "gene:braf",
      "name": "BRAF",
      "entity_type": "gene",
      "confidence": 0.98
    },
    {
      "id": "gene:map2k1",
      "name": "MEK1",
      "entity_type": "gene",
      "confidence": 0.97
    },
    {
      "id": "gene:map2k2",
      "name": "MEK2",
      "entity_type": "gene",
      "confidence": 0.96
    },
    {
      "id": "pathway:mapk_signaling",
      "name": "MAPK signaling pathway",
      "entity_type": "pathway",
      "confidence": 0.99
    },
    {
      "id": "drug:vemurafenib",
      "name": "vemurafenib",
      "entity_type": "drug",
      "confidence": 0.96
    },
    {
      "id": "drug:dabrafenib",
      "name": "dabrafenib",
      "entity_type": "drug",
      "confidence": 0.95
    },
    {
      "id": "disease:melanoma",
      "name": "melanoma",
      "entity_type": "disease",
      "confidence": 0.97
    },
    {
      "id": "mutation:braf_v600e",
      "name": "BRAF_V600E",
      "entity_type": "mutation",
      "confidence": 0.98
    },
    {
      "id": "gene:mapk1",
      "name": "ERK2",
      "entity_type": "gene",
      "confidence": 0.95
    },
    {
      "id": "drug:trametinib",
      "name": "trametinib",
      "entity_type": "drug",
      "confidence": 0.94
    }
  ],
  "edges": [
    {
      "source_id": "gene:braf",
      "target_id": "gene:map2k1",
      "relationship_type": "phosphorylates",
      "confidence": 0.96,
      "pch_layer": 2
    },
    {
      "source_id": "gene:braf",
      "target_id": "gene:map2k2",
      "relationship_type": "phosphorylates",
      "confidence": 0.94,
      "pch_layer": 2
    },
    {
      "source_id": "gene:braf",
      "target_id": "pathway:mapk_signaling",
      "relationship_type": "member_of",
      "confidence": 0.99,
      "pch_layer": 1
    },
    {
      "source_id": "drug:vemurafenib",
      "target_id": "gene:braf",
      "relationship_type": "targets",
      "confidence": 0.96,
      "pch_layer": 2
    },
    {
      "source_id": "drug:dabrafenib",
      "target_id": "gene:braf",
      "relationship_type": "targets",
      "confidence": 0.95,
      "pch_layer": 2
    },
    {
      "source_id": "gene:braf",
      "target_id": "disease:melanoma",
      "relationship_type": "driver_in",
      "confidence": 0.97,
      "pch_layer": 2
    },
    {
      "source_id": "mutation:braf_v600e",
      "target_id": "gene:braf",
      "relationship_type": "variant_in_gene",
      "confidence": 0.98,
      "pch_layer": 1
    },
    {
      "source_id": "gene:map2k1",
      "target_id": "gene:mapk1",
      "relationship_type": "phosphorylates",
      "confidence": 0.95,
      "pch_layer": 2
    },
    {
      "source_id": "drug:trametinib",
      "target_id": "gene:map2k1",
      "relationship_type": "targets",
      "confidence": 0.94,
      "pch_layer": 2
    },
    {
      "source_id": "gene:map2k1",
      "target_id": "pathway:mapk_signaling",
      "relationship_type": "member_of",
      "confidence": 0.98,
      "pch_layer": 1
    }
  ],
  "node_count": 10,
  "edge_count": 10
}

Try It

API KeyGet your free API key →

Path Parameters

namestringrequired

Entity name to build subgraph around

Query Parameters

radiusinteger

Number of hops from center entity (1–3)

max_nodesinteger

Maximum nodes to include (10–1000)

GET https://research.usegalen.com/api/v1/entities/{name}/subgraph

Related Endpoints

Get Entity Profile Get Entity Relationships Find Paths Between Entities

100–500ms

GET/knowledge/pch

Free

Get PCH Annotations

Get the knowledge-graph-wide Pearl Causal Hierarchy distribution.

Only in Galen

The Pearl Causal Hierarchy is Galen's foundational framework. Every relationship is classified as L1 (seeing — statistical association), L2 (doing — experimental intervention), or L3 (imagining — counterfactual reasoning). This endpoint gives you the exact evidence quality distribution for any entity. No other cancer API applies formal causal epistemology.

Response Schema

total_entitiesintegerTotal entities in the knowledge graph

l1_countintegerNumber of L1 (associational) relationships

l2_countintegerNumber of L2 (interventional) relationships — from CRISPR, drug screens, etc.

l3_countintegerNumber of L3 (counterfactual) relationships — validated via Structural Causal Models

l1_fractionfloatFraction of relationships at L1

l2_fractionfloatFraction of relationships at L2

l3_fractionfloatFraction of relationships at L3

totalintegerTotal number of relationships

Example Request

import requests

resp = requests.get(
    "https://research.usegalen.com/api/v1/knowledge/pch",
    headers={"X-API-Key": "YOUR_API_KEY"},
)
pch = resp.json()
print(f"KG-wide PCH distribution:")
print(f"  L1 (associations):    {pch['l1_count']} ({pch['l1_fraction']:.0%})")
print(f"  L2 (interventions):   {pch['l2_count']} ({pch['l2_fraction']:.0%})")
print(f"  L3 (counterfactuals): {pch['l3_count']} ({pch['l3_fraction']:.0%})")

Example Response

{
  "total_entities": 283421,
  "l1_count": 3247891,
  "l2_count": 1287432,
  "l3_count": 215960,
  "l1_fraction": 0.683,
  "l2_fraction": 0.271,
  "l3_fraction": 0.045,
  "total": 4751283
}

Try It

API KeyGet your free API key →

GET https://research.usegalen.com/api/v1/knowledge/pch

Related Endpoints

Get Entity Relationships Get Entity Profile

20–100ms

GET/relationships/paths

Free

Find Paths Between Entities

Discover all paths connecting two entities in the knowledge graph.

Only in Galen

Discover how any two cancer biology entities are connected through the knowledge graph. Paths traverse typed, directed edges with causal annotations — showing you the biological mechanism chain, not just that a connection exists.

Parameters

Query Parameters

Name	Type	Description
`source`	string	Starting entity Example: `EGFR`
`target`	string	Destination entity Example: `erlotinib`
`max_hops`	integer	Maximum path length (1–5) Default: `3`
`max_paths`	integer	Maximum number of paths to return (1–50) Default: `10`

Response Schema

sourcestringThe source entity identifier

targetstringThe target entity identifier

pathsarrayArray of paths, each an array of hop objects

├──entitystringEntity at this position in the path

├──relationship_typestringRelationship type traversed

└──directionstringEdge direction (forward or reverse)

path_countintegerNumber of paths found

Example Request

import requests

resp = requests.get(
    "https://research.usegalen.com/api/v1/relationships/paths",
    params={"source": "EGFR", "target": "erlotinib", "max_hops": 3, "max_paths": 10},
    headers={"X-API-Key": "YOUR_API_KEY"},
)
data = resp.json()
print(f"Found {data['path_count']} paths from {data['source']} to {data['target']}:")
for i, path in enumerate(data["paths"], 1):
    steps = " → ".join(step["entity"] for step in path)
    print(f"  Path {i}: {steps}")

Example Response

{
  "source": "gene:egfr",
  "target": "drug:erlotinib",
  "paths": [
    [
      {
        "entity": "EGFR",
        "relationship_type": "targets",
        "direction": "reverse"
      },
      {
        "entity": "erlotinib",
        "relationship_type": "targets",
        "direction": "forward"
      }
    ],
    [
      {
        "entity": "EGFR",
        "relationship_type": "member_of",
        "direction": "forward"
      },
      {
        "entity": "MAPK signaling pathway",
        "relationship_type": "member_of",
        "direction": "reverse"
      },
      {
        "entity": "erlotinib",
        "relationship_type": "targets",
        "direction": "forward"
      }
    ],
    [
      {
        "entity": "EGFR",
        "relationship_type": "driver_in",
        "direction": "forward"
      },
      {
        "entity": "NSCLC",
        "relationship_type": "indicated_for",
        "direction": "reverse"
      },
      {
        "entity": "erlotinib",
        "relationship_type": "indicated_for",
        "direction": "forward"
      }
    ]
  ],
  "path_count": 3
}

Try It

API KeyGet your free API key →

Query Parameters

sourcestringrequired

Starting entity

targetstringrequired

Destination entity

max_hopsinteger

Maximum path length (1–5)

max_pathsinteger

Maximum number of paths to return (1–50)

GET https://research.usegalen.com/api/v1/relationships/paths

Related Endpoints

Get Entity Relationships Extract Subgraph Explain a Causal Relationship

200ms–5s