tooluniverse-gwas-drug-discovery — Claude Code skill: how to install

---
name: tooluniverse-gwas-drug-discovery
description: Transform GWAS signals into drug targets and repurposing opportunities. Connects GWAS-significant loci to causal genes via fine-mapping/eQTL, then to druggable proteins via DGIdb/OpenTargets, then to existing drugs via ChEMBL. Use for GWAS-to-target hypothesis generation, druggable-fraction analysis of disease loci, and human-genetics-validated drug-repurposing prioritization.
disable-model-invocation: true
---

# GWAS-to-Drug Target Discovery

Transform genome-wide association studies (GWAS) into actionable drug targets and repurposing opportunities.

**IMPORTANT**: Always use English terms in tool calls. Respond in the user's language.

---

## Overview

This skill bridges genetic discoveries from GWAS with drug development by:

1. **Identifying genetic risk factors** - Finding genes associated with diseases
2. **Assessing druggability** - Evaluating which genes can be targeted by drugs
3. **Prioritizing targets** - Ranking candidates by genetic evidence strength
4. **Finding existing drugs** - Discovering approved/investigational compounds
5. **Identifying repurposing opportunities** - Matching drugs to new indications

**Key insight**: Targets with genetic support have 2x higher probability of clinical approval (Nelson et al., Nature Genetics 2015).

## Reasoning Strategy

GWAS-to-drug translation succeeds when you think causally. A genetic association provides causal direction that observational data cannot: if a loss-of-function variant protects against disease, an inhibitor of that gene's product is the hypothesis to test. The direction of effect (LOF vs. GOF) determines whether you need an inhibitor or an agonist — get this wrong and the drug works backwards. GWAS effect sizes are small (odds ratios of 1.1–1.5 are typical), but the drug effect may be much larger or smaller than the genetic effect; the genetic signal validates the target, not the dose. Always integrate multiple lines of evidence (eQTL colocalization, pQTL, L2G score) before committing to a target, because many GWAS variants tag the causal gene only indirectly.

**LOOK UP DON'T GUESS**: Do not assume which gene a GWAS variant implicates — use `OpenTargets_get_variant_credible_sets` or `gwas_get_associations_for_snp` to get the actual mapped gene and L2G score. Do not guess the direction of effect, odds ratio, or whether a drug already exists for the target; always query the tools.

---

## Workflow Steps

### Step 1: GWAS Gene Discovery

**Input**: Disease/trait name (e.g., "type 2 diabetes", "Alzheimer disease")

**Process**: Query GWAS Catalog for associations, filter by significance (p < 5x10^-8), map variants to genes, aggregate evidence.

**Tools**:
- `gwas_get_associations_for_trait` - Get associations by disease
- `gwas_search_associations` - Flexible search
- `gwas_get_associations_for_snp` - SNP-specific associations
- `OpenTargets_search_gwas_studies_by_disease` - Curated GWAS data
- `OpenTargets_get_variant_credible_sets` - Fine-mapped loci with L2G predictions

### Step 2: Druggability Assessment

**Input**: Gene list from Step 1

**Process**: Check target class, assess tractability, evaluate safety, check for tool compounds or structures.

**Tools**:
- `OpenTargets_get_target_tractability_by_ensemblID` - Druggability assessment
- `OpenTargets_get_target_classes_by_ensemblID` - Target classification
- `OpenTargets_get_target_safety_profile_by_ensemblID` - Safety data
- `OpenTargets_get_target_genomic_location_by_ensemblID` - Genomic context

### Step 3: Target Prioritization

**Scoring Formula**:
```
Target Score = (GWAS Score x 0.4) + (Druggability x 0.3) + (Clinical Evidence x 0.2) + (Novelty x 0.1)
```

Rank targets by composite score. Generate target dossiers.

### Step 4: Existing Drug Search

**Process**: Search drug-target associations, find approved drugs and clinical candidates, get MOA and indication data.

**Tools**:
- `OpenTargets_get_associated_drugs_by_disease_efoId` - Known drugs for disease
- `OpenTargets_get_drug_mechanisms_of_action_by_chemblId` - Drug MOA
- `ChEMBL_get_target_activities` - Bioactivity data
- `ChEMBL_get_drug_mechanisms` / `ChEMBL_search_drugs` - Drug data

### Step 5: Clinical Evidence & Safety

**Tools**:
- `FDA_get_adverse_reactions_by_drug_name` - Safety data
- `FDA_get_active_ingredient_info_by_drug_name` - Drug composition
- `OpenTargets_get_drug_warnings_by_chemblId` - Drug warnings

### Step 6: Repurposing Opportunities

Match drug targets to new disease genes, assess mechanistic fit, check contraindications, estimate repurposing probability.

---

## Quick Start

```python
from tooluniverse import ToolUniverse
tu = ToolUniverse()
tu.load_tools()

# Step 1: Get GWAS associations (use disease_trait not trait; no p_value_threshold param)
associations = tu.tools.gwas_get_associations_for_trait(disease_trait="type 2 diabetes")

# Step 2: Assess druggability (ensemblId lowercase d)
tractability = tu.tools.OpenTargets_get_target_tractability_by_ensemblID(ensemblId="ENSG00000148737")

# Step 3: Find existing drugs per target via DGIdb (OpenTargets drug query may return HTTP 400)
drugs = tu.tools.DGIdb_get_drug_gene_interactions(genes=["TCF7L2"])
```

---

## All Tools by Category

**GWAS & Genetics**:
- `gwas_get_associations_for_trait` / `gwas_search_associations` / `gwas_get_associations_for_snp`
- `OpenTargets_search_gwas_studies_by_disease` / `OpenTargets_get_variant_credible_sets`

**Target Assessment**:
- `OpenTargets_get_target_tractability_by_ensemblID` / `OpenTargets_get_target_classes_by_ensemblID`
- `OpenTargets_get_target_safety_profile_by_ensemblID` / `OpenTargets_get_target_genomic_location_by_ensemblID`

**Drug Discovery**:
- `OpenTargets_get_associated_drugs_by_disease_efoId` / `OpenTargets_get_drug_mechanisms_of_action_by_chemblId`
- `ChEMBL_get_target_activities` / `ChEMBL_get_drug_mechanisms` / `ChEMBL_search_drugs`

**Safety & Clinical**:
- `FDA_get_adverse_reactions_by_drug_name` / `FDA_get_active_ingredient_info_by_drug_name`
- `OpenTargets_get_drug_warnings_by_chemblId`

**Literature**:
- `PubMed_search_articles` / `EuropePMC_search_articles` / `ClinicalTrials_search_studies`

---

## Best Practices

1. **Multi-ancestry GWAS**: Include trans-ethnic meta-analyses for robust signals
2. **Functional validation**: Confirm with eQTL, pQTL, colocalization analysis
3. **Network analysis**: Group GWAS hits by pathway (KEGG, Reactome)
4. **Safety assessment**: Check gnomAD pLI, GTEx expression, PharmaGKB
5. **Batch operations**: Use `tu.run_batch()` for parallel queries across targets

---

## Parameter Gotchas

| Issue | Wrong | Correct |
|-------|-------|---------|
| GWAS trait param | `gwas_get_associations_for_trait(trait=...)` | `disease_trait=...` (no `trait` param exists) |
| GWAS p-value filter | `p_value_threshold=5e-8` | No such param; filter client-side after fetching results |
| OpenTargets ensembl case | `ensemblID="ENSG..."` | `ensemblId="ENSG..."` (lowercase 'd') |
| ClinicalTrials tool name | `ClinicalTrials_search_studies(...)` | `ClinicalTrials_search_studies(...)` |
| DGIdb tool name | `DGIdb_get_drug_gene_interactions(...)` | `DGIdb_get_drug_gene_interactions(genes=[...])` |
| OpenTargets disease drugs | `OpenTargets_get_associated_drugs_by_disease_efoId` may return HTTP 400 | Fall back to `DGIdb_get_drug_gene_interactions` per gene |
| GWAS study search param | `gwas_search_studies(disease_trait=...)` | Use `efo_trait=...` for studies (disease_trait works for associations only) |

## Interpretation: From GWAS Hit to Drug Target

### GWAS Signal Strength Assessment

| Signal Quality | Criteria | Drug Discovery Value |
|---------------|---------|---------------------|
| **Gold standard** | Genome-wide significant (p < 5e-8), replicated across ancestries, L2G > 0.5, eQTL colocalized | Highest priority — genetic causality established |
| **Strong** | Genome-wide significant, L2G > 0.3, biological plausibility | High priority — pursue with functional validation |
| **Moderate** | Suggestive (p < 1e-5), or significant but no fine-mapping | Medium — needs additional evidence before investment |
| **Weak** | Single study, no replication, low L2G, no functional support | Low — hypothesis generating only |

### Target Prioritization Decision Tree

After identifying GWAS-linked genes, rank them by answering:

1. **Is the gene druggable?** (DGIdb category: kinase/GPCR/ion channel = yes; transcription factor/scaffold = harder)
   - If approved drug exists → **REPURPOSING** opportunity (fastest path)
   - If druggable but no drug → **NOVEL TARGET** (standard drug discovery)
   - If not druggable → consider antisense/PROTAC/genetic medicine

2. **Is the genetic direction clear?**
   - LOF variants increase disease risk → need an AGONIST or gene therapy
   - GOF variants increase disease risk → need an INHIBITOR (typical small molecule)
   - Direction unclear → need functional studies before drug design

3. **What's the effect size?** (Odds ratio from GWAS)
   - OR > 2.0: strong effect, likely penetrant → Mendelian-like, high confidence
   - OR 1.2-2.0: moderate, common in complex disease → validate with independent data
   - OR < 1.2: small effect → may not be clinically meaningful alone

4. **Is there clinical precedent?**
   - Drug for same target approved for ANY disease → safety data exists → lower risk
   - Drug in clinical trials → partial de-risking
   - No precedent → full de novo development risk

---

## Troubleshooting

| Problem | Solution |
|---------|----------|
| No GWAS hits for disease | Try broader trait name, check synonyms, use OpenTargets |
| Gene not in druggable class | Consider antibody/antisense modalities, check pathway neighbors |
| No existing drugs for target | Target may be novel - check tool compounds in ChEMBL |
| Low L2G score | Variants may be regulatory - check eQTL/pQTL evidence |

---

## Reference Files

- **REFERENCE.md** - Detailed concepts, druggability tiers, clinical translation, limitations, ethics
- **EXAMPLES.md** - Use cases (Huntington's, Alzheimer's, diabetes) with success stories
- **REPORT_TEMPLATE.md** - Output report template with scoring criteria
- **PROCEDURES.md** - Step-by-step implementation procedures
- **QUICK_START.md** - Quick start guide
- Related skills: tooluniverse-drug-repurposing, disease-intelligence-gatherer, tooluniverse-sdk