POST
/patient/profile/interpretPro
Interpret Patient Mutation Profile
Interpret a patient's multi-mutation cancer profile with evidence-graded annotations and causal analysis.
Only in Galen
Multi-mutation interpretation grounded in experimental evidence from 10+ databases — not an LLM summarizing literature. Every annotation is traced to its source with a Pearl Causal Hierarchy layer. Evidence quality is graded by the fraction of L2+ (interventional and counterfactual) relationships used. Mutation interactions are identified from real co-occurrence data across 498 clinical studies.
Parameters
Request Body
Patient mutation profile including cancer type, mutations, and optional clinical context.
| Field | Type | Description |
|---|---|---|
cancer_type | string | Cancer type or subtype |
mutations | array | List of mutations in GENE VARIANT format (e.g. EGFR L858R) |
stage | string | Cancer stage (optional) |
prior_treatments | array | Previous treatments received (optional) |
Response Schema
cancer_typestringThe cancer type analyzedvariant_annotationsarrayPer-mutation annotations with oncogenic effect, evidence level, and causal hierarchy layer├──
genestringGene symbol├──
variantstringVariant designation├──
oncogenic_effectstringFunctional effect — gain-of-function, loss-of-function, etc.├──
evidence_levelstringESCAT clinical evidence tier├──
actionabilitystringWhether this mutation is actionable (e.g. FDA-approved therapy exists)├──
pathogenicitystringPathogenicity classification (pathogenic, likely pathogenic, VUS, etc.)├──
frequencyfloatMutation frequency in the specified cancer type (0.0–1.0)├──
pch_layerintegerHighest Pearl Causal Hierarchy layer of evidence (1, 2, or 3)└──
confidencefloatAnnotation confidence score (0.0–1.0)mutation_interactionsarrayPairwise interactions between mutations in the profile├──
gene_astringFirst gene├──
gene_bstringSecond gene├──
interaction_typestringType of interaction (e.g. co-occurring, synthetic_lethal, epistatic)├──
clinical_significancestringClinical relevance of this interaction└──
co_occurrence_frequencyfloatHow often these mutations co-occur (0.0–1.0)causal_summarystring | nullNarrative summary of causal relationships between mutationsnarrativestring | nullPatient-friendly explanation of the mutation profileevidence_qualityobjectAssessment of the evidence quality backing this interpretation├──
gradestringOverall evidence grade: high, moderate, low, or insufficient├──
l2_l3_fractionfloatFraction of evidence from experimental (L2) and counterfactual (L3) sources├──
grounded_sourcesarrayData sources that contributed to this interpretation├──
source_namestringName of the data source└──
relationship_countintegerNumber of relationships from this source used in the interpretation└──
ungrounded_claim_countintegerNumber of claims that could not be grounded in source datadisclaimerstringMedical disclaimerExample Request
import requests
resp = requests.post(
"https://research.usegalen.com/api/v1/patient/profile/interpret",
headers={"X-API-Key": "YOUR_API_KEY"},
json={
"cancer_type": "NSCLC",
"mutations": ["EGFR L858R", "TP53 R248W"],
"stage": "IV",
"prior_treatments": ["carboplatin", "pemetrexed"],
},
)
profile = resp.json()
for v in profile["variant_annotations"]:
print(f"{v['gene']} {v['variant']}: {v['oncogenic_effect']} (evidence: {v['evidence_level']})")
print(f"Evidence quality: {profile['evidence_quality']['grade']}")Example Response
{
"cancer_type": "NSCLC",
"variant_annotations": [
{
"gene": "EGFR",
"variant": "L858R",
"oncogenic_effect": "gain-of-function",
"evidence_level": "Level 1",
"actionability": "Actionable — FDA-approved therapies (erlotinib, osimertinib, afatinib)",
"pathogenicity": "pathogenic",
"frequency": 0.38,
"pch_layer": 2,
"confidence": 0.96
},
{
"gene": "TP53",
"variant": "R248W",
"oncogenic_effect": "loss-of-function",
"evidence_level": "Level 3B",
"actionability": "Not directly actionable — no FDA-approved TP53-targeted therapy",
"pathogenicity": "pathogenic",
"frequency": 0.07,
"pch_layer": 2,
"confidence": 0.93
}
],
"mutation_interactions": [
{
"gene_a": "EGFR",
"gene_b": "TP53",
"interaction_type": "co-occurring",
"clinical_significance": "TP53 co-mutation associated with shorter PFS on EGFR TKI therapy in NSCLC. Median PFS reduction of ~3.5 months compared to EGFR-mutant/TP53-wildtype.",
"co_occurrence_frequency": 0.54
}
],
"causal_summary": "EGFR L858R constitutively activates the EGFR kinase domain, driving proliferation through RAS-MAPK and PI3K-AKT signaling. TP53 R248W disrupts DNA-binding and eliminates tumor suppressor function. The co-occurrence of these mutations suggests genomic instability (via TP53 loss) compounding oncogenic signaling (via EGFR activation), which is associated with inferior outcomes on EGFR TKI monotherapy.",
"narrative": "Your tumor has two important mutations. The EGFR L858R mutation is a well-known driver of lung cancer growth, and the good news is that there are several FDA-approved targeted therapies available for it, including osimertinib. The TP53 R248W mutation affects a key tumor suppressor gene. When both mutations are present together, some studies suggest that targeted therapy alone may be less effective than expected. Your oncologist may want to discuss combination strategies or clinical trial options.",
"evidence_quality": {
"grade": "moderate",
"l2_l3_fraction": 0.65,
"grounded_sources": [
{
"source_name": "cbioportal",
"relationship_count": 47
},
{
"source_name": "chembl_36",
"relationship_count": 31
},
{
"source_name": "cosmic",
"relationship_count": 28
},
{
"source_name": "oncokb",
"relationship_count": 12
},
{
"source_name": "intogen",
"relationship_count": 8
}
],
"ungrounded_claim_count": 0
},
"disclaimer": "This information is for research and educational purposes only. It does not constitute medical advice, diagnosis, or treatment. The Galen API is not FDA-approved for clinical decision-making. Always consult a qualified healthcare provider for medical decisions. Galen Health is not responsible for actions taken based on this information."
}Try It
Request Body
Patient mutation profile including cancer type, mutations, and optional clinical context.
POST https://research.usegalen.com/api/v1/patient/profile/interpret5–30s
This information is for research and educational purposes only. It does not constitute medical advice, diagnosis, or treatment. The Galen API is not FDA-approved for clinical decision-making. Always consult a qualified healthcare provider for medical decisions. Galen Health is not responsible for actions taken based on this information.