SOP: finding druggability and target-disease evidence

Use this SOP when assigning druggability-tier: to protein, pathway, or process pages, or when researching whether a target has clinical-stage drugs or high-quality chemical probes. The primary resources are Open Targets Platform (disease-target evidence aggregation), Open Targets Genetics (GWAS-to-gene), DGIdb (drug-gene interactions), and PROTACDB (targeted protein degradation).

When to use druggability resources

Filling the druggability-tier: field on type: protein, type: pathway, or type: process pages.
Checking whether a target has clinical-stage compounds — to populate clinical-trials-active: on compound pages or to write a “Clinical evidence” section.
Assessing whether a target is genetically validated (GWAS or rare-variant evidence linking it to an aging-relevant disease).
Finding the best chemical probe for a target when no drug exists yet (tier 2).

Resource 1: Open Targets Platform

What it is: Open Targets Platform aggregates evidence linking targets (genes) to diseases from genetics, drugs, RNA expression, animal models, and text mining. The most comprehensive disease–target evidence database.

Primary endpoint: https://api.platform.opentargets.org/api/v4/graphql

Method: POST with GraphQL query body.

Query: target overview (druggability + tractability)

curl -X POST "https://api.platform.opentargets.org/api/v4/graphql" \
  -H "Content-Type: application/json" \
  -d '{
    "query": "{ target(ensemblId: \"ENSG00000141510\") { id approvedSymbol tractability { label modality value } drugAndClinicalCandidates { count rows { maxClinicalStage drug { name id drugType maximumClinicalStage } } } } }"
  }'

Key fields in tractability response:

modality — SM (small molecule), AB (antibody), OC (other clinical)
label — e.g., Clinical Precedence, Discovery Precedence, Predicted Tractable
value — boolean (is there evidence for this modality?)

Schema note (verified 2026-05-06): The knownDrugs field on the Target type referenced in earlier versions of this SOP no longer exists. It has been replaced by drugAndClinicalCandidates, which returns ClinicalTargetFromTarget rows with a slightly different shape — drug metadata is nested under a drug sub-object rather than flat on the row. If you encounter Cannot query field 'knownDrugs' on type 'Target', you’re hitting the old query.

Druggability tier mapping (R14 schema)

`druggability-tier:` value	Open Targets tractability signal
`1`	SM or AB `Clinical Precedence = true` (approved or clinical-stage drug exists)
`2`	SM or AB `Discovery Precedence = true` (high-quality probe; no approved drug yet)
`3`	SM `Predicted Tractable` only (ChEMBL structure-based prediction)
`4`	No tractability evidence; no pocket identified; or actively “undruggable” class

When there is no Open Targets entry for a gene (ncRNA, very small ORF, etc.), set druggability-tier: null and tag #gap/no-opentargets-entry.

Query: target-disease association score

curl -X POST "https://api.platform.opentargets.org/api/v4/graphql" \
  -H "Content-Type: application/json" \
  -d '{
    "query": "{ target(ensemblId: \"ENSG00000141510\") { associatedDiseases(page: {index: 0, size: 5}) { rows { disease { name id } score datasourceScores { id score } } } } }"
  }'

Use EFO IDs for aging-relevant disease searches: Alzheimer’s (EFO_0000249), type 2 diabetes (EFO_0001360), sarcopenia (EFO_0005775), frailty (HP:0003077), cardiovascular disease (EFO_0000319).

Query: list known drugs for a target

curl -X POST "https://api.platform.opentargets.org/api/v4/graphql" \
  -H "Content-Type: application/json" \
  -d '{
    "query": "{ target(ensemblId: \"ENSG00000141510\") { drugAndClinicalCandidates { count rows { maxClinicalStage drug { name id drugType maximumClinicalStage } } } } }"
  }'

Phase values (current enum): PHASE_1, PHASE_1_2, PHASE_2, PHASE_3, PHASE_4, APPROVAL — string enum, not numeric. APPROVAL indicates an approved drug. Record the highest stage reached.

To get mechanism-of-action and indication detail per drug, query the Drug type separately using the returned drug.id (ChEMBL ID):

curl -X POST "https://api.platform.opentargets.org/api/v4/graphql" \
  -H "Content-Type: application/json" \
  -d '{
    "query": "{ drug(chemblId: \"CHEMBL1421\") { name maximumClinicalStage drugType tradeNames mechanismsOfAction { rows { mechanismOfAction targetName actionType } } indications { rows { disease { name id } maxClinicalStage } } drugWarnings { warningType description } } }"
  }'

Useful Drug fields: name, tradeNames, drugType, maximumClinicalStage, mechanismsOfAction, indications, drugWarnings, adverseEvents, synonyms. The ClinicalIndicationFromDrug row exposes per-indication phase as maxClinicalStage (not maxPhaseForIndication or phase). The ClinicalTargetFromTarget row also carries clinicalReports and per-target diseases if you want indication context without a second query.

Resource 2: Open Targets Genetics

What it is: Open Targets Genetics maps GWAS loci to likely causal genes using Locus-to-Gene (L2G) scoring, integrating QTL data, chromatin accessibility, and distance. Useful for genetic validation of targets.

Primary endpoint: https://api.genetics.opentargets.org/graphql

Query: L2G scores for a GWAS study

curl -X POST "https://api.genetics.opentargets.org/graphql" \
  -H "Content-Type: application/json" \
  -d '{
    "query": "{ studyLocus2GeneTable(studyId: \"GCST90012130\", variantId: \"13_33628138_T_A\") { rows { gene { id symbol } score } } }"
  }'

Query: find GWAS studies for a target gene

curl -X POST "https://api.genetics.opentargets.org/graphql" \
  -H "Content-Type: application/json" \
  -d '{
    "query": "{ geneInfo(geneId: \"ENSG00000141510\") { id symbol } }"
  }'

L2G score interpretation: Score ≥0.5 = reasonable confidence the GWAS locus is driven by variation in this gene. Score ≥0.8 = high confidence.

Resource 3: DGIdb (Drug Gene Interaction Database)

What it is: DGIdb aggregates drug-gene interaction data from >40 source databases (ChEMBL, DrugBank, TTD, PharmGKB, etc.). Covers small molecules, biologics, and nutraceuticals. Maintained by Washington University.

Current API version: v5 (GraphQL). The v2 REST API (shown in some older references) is deprecated; WebFetch of v2 endpoints returns only the DGIdb homepage rather than JSON data.

Primary GraphQL endpoint: https://dgidb.org/graphql

# Query drug interactions for a gene
curl -X POST "https://dgidb.org/graphql" \
  -H "Content-Type: application/json" \
  -d '{
    "query": "{ genes(names: [\"MTOR\"]) { nodes { name interactions { drug { name conceptId } interactionScore interactionTypes { type directionality } sources { sourceDbName } } } } }"
  }'

Key fields:

drug.name — drug name
drug.conceptId — ChEMBL or DrugBank ID
interactionScore — aggregated confidence score
interactionTypes[].type — e.g., inhibitor, activator, agonist
interactionTypes[].directionality — inhibitory | activating
sources[].sourceDbName — which databases support this interaction

Note on WebFetch: DGIdb serves a React SPA; the v2 REST endpoints return only the HTML shell when accessed without the full JS runtime. Use the GraphQL endpoint via curl/Bash or the Python dgidb package instead.

Python package:

from dgidb import get_interactions
result = get_interactions(genes=["MTOR", "TP53"])

Resource 4: PROTACDB

What it is: A database of PROTAC (Proteolysis Targeting Chimera) molecules — targeted protein degraders. Relevant for proteins considered “undruggable” by conventional small molecules but tractable via degradation.

Repository: https://github.com/Wang-Lin-boop/PROTACDB

Access: No live REST API. Data available as downloadable CSV/JSON from the GitHub repository releases. Query locally.

When to use: When a target has druggability-tier: 4 by conventional criteria but is biologically important, check PROTACDB for PROTAC development status. Note on the page as: “PROTAC development reported but no clinical-stage compound” if found.

Populating `druggability-tier:` — decision tree

1. Does Open Targets have an entry for this gene?
   NO  → druggability-tier: null + #gap/no-opentargets-entry
   YES → continue

2. Is Clinical Precedence (SM or AB) = true?
   YES → druggability-tier: 1
   NO  → continue

3. Is Discovery Precedence (SM or AB) = true?
   YES → druggability-tier: 2
   NO  → continue

4. Is Predicted Tractable (SM) = true?
   YES → druggability-tier: 3
   NO  → continue

5. Is there a PROTAC in PROTACDB?
   YES → druggability-tier: 4 (note: PROTAC reported; not yet clinical)
   NO  → druggability-tier: 4

Recording clinical advance data (`targeted-clinical-trials:`)

This optional field on protein pages (not a current frontmatter standard, but useful in the body) counts how many clinical trials target this gene directly. Pull from the drugAndClinicalCandidates.count in the Open Targets Platform query above — this count already represents drugs that have reached at least Phase 1 (preclinical-only entries are not included in drugAndClinicalCandidates).

For compound and intervention pages, use the clinical-trials-active: field instead — see sops/integrating-clinical-trials.md.

Workflow for a protein page

Resolve the ENSG ID from UniProt or GTEx reference.
Run Open Targets Platform tractability query → assign druggability-tier:.
Record the highest-phase drug(s) in the body’s “Known modulators” section.
Run DGIdb GraphQL query → add any additional inhibitors/activators not in Open Targets.
Check Open Targets Genetics L2G for genetic validation of target against aging-relevant diseases.
Note PROTAC status from PROTACDB if tier = 4.

🧬 Aging Wiki

Explorer

finding-druggability

SOP: finding druggability and target-disease evidence

When to use druggability resources

Resource 1: Open Targets Platform

Query: target overview (druggability + tractability)

Druggability tier mapping (R14 schema)

Query: target-disease association score

Query: list known drugs for a target

Resource 2: Open Targets Genetics

Query: L2G scores for a GWAS study

Query: find GWAS studies for a target gene

Resource 3: DGIdb (Drug Gene Interaction Database)

Resource 4: PROTACDB

Populating `druggability-tier:` — decision tree

Recording clinical advance data (`targeted-clinical-trials:`)

Workflow for a protein page

See also

Graph View

Table of Contents

Backlinks

🧬 Aging Wiki

Explorer

finding-druggability

SOP: finding druggability and target-disease evidence

When to use druggability resources

Resource 1: Open Targets Platform

Query: target overview (druggability + tractability)

Druggability tier mapping (R14 schema)

Query: target-disease association score

Query: list known drugs for a target

Resource 2: Open Targets Genetics

Query: L2G scores for a GWAS study

Query: find GWAS studies for a target gene

Resource 3: DGIdb (Drug Gene Interaction Database)

Resource 4: PROTACDB

Populating druggability-tier: — decision tree

Recording clinical advance data (targeted-clinical-trials:)

Workflow for a protein page

See also

Graph View

Table of Contents

Backlinks

Populating `druggability-tier:` — decision tree

Recording clinical advance data (`targeted-clinical-trials:`)