tooluniverse-drug-repurposing

Drug Repurposing with ToolUniverse

Systematically identify and evaluate drug repurposing candidates using multiple computational strategies.

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

Reasoning Before Searching

Start by asking: WHY might this drug work for a new disease? Three strategies:

• (a) Same target: The drug's primary target is also involved in the new disease. This is the strongest hypothesis — use OpenTargets to check if the target has genetic evidence in both diseases before any other search.
• (b) Off-target activity: The drug has secondary targets or off-target effects that are relevant to the new disease. Check ChEMBL bioactivity data for all known targets of the drug, not just its primary one.
• (c) Shared pathways: The original indication and new disease share molecular pathways, even if the target itself is not genetically linked. Use Reactome and STRING to compare pathway overlap between diseases.

Each strategy uses different tools and has different evidentiary weight. Identify which strategy applies FIRST, then choose the corresponding workflow below. Do not run all three strategies blindly — reason about which is most plausible given the drug's mechanism.

LOOK UP DON'T GUESS: Never assume a drug hits a target, never assume a target is disease-relevant, never assume pathway overlap. Verify each link with tool calls.

Core Strategies

1. Target-Based: Disease targets -> Find drugs that modulate those targets
2. Compound-Based: Approved drugs -> Find new disease indications
3. Disease-Driven: Disease -> Targets -> Match to existing drugs

Workflow Overview

Phase 1: Disease & Target Analysis
  Get disease info (OpenTargets), find associated targets, get target details

Phase 2: Drug Discovery
  Search DrugBank, DGIdb, ChEMBL for drugs targeting disease-associated genes
  Get drug details, indications, pharmacology

Phase 3: Safety & Feasibility Assessment
  FDA warnings, FAERS adverse events, drug interactions, ADMET predictions

Phase 4: Literature Evidence
  PubMed, Europe PMC, clinical trials for existing evidence

Phase 5: Scoring & Ranking
  Composite score: target association + safety + literature + drug properties

See: PROCEDURES.md for detailed step-by-step procedures and code patterns.

Quick Start

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

# Step 1: Get disease targets
disease_info = tu.tools.OpenTargets_get_disease_id_description_by_name(diseaseName="rheumatoid arthritis")
# Response nests ID at data.search.hits[0].id
disease_id = disease_info['data']['search']['hits'][0]['id']
targets = tu.tools.OpenTargets_get_associated_targets_by_disease_efoId(efoId=disease_id, limit=10)

# Step 2: Find drugs for each target
# Response nests targets at data.disease.associatedTargets.rows
rows = targets['data']['disease']['associatedTargets']['rows']
for target in rows[:5]:
    gene = target['target']['approvedSymbol']
    drugs = tu.tools.DGIdb_get_drug_gene_interactions(genes=[gene])

Key ToolUniverse Tools

Disease & Target:

• OpenTargets_get_disease_id_description_by_name - Disease lookup
• OpenTargets_get_associated_targets_by_disease_efoId - Disease targets
• UniProt_get_entry_by_accession - Protein details

Drug Discovery:

• drugbank_get_drug_name_and_description_by_target_name - Drugs by target. Param: query= (NOT target_name=)
• drugbank_get_drug_name_and_description_by_indication - Drugs by indication. Param: query= (NOT indication=)
• DGIdb_get_drug_gene_interactions - Drug-gene interactions. Response path: data.data.genes.nodes[0].interactions
• ChEMBL_search_drugs / ChEMBL_get_drug_mechanisms - Drug search and MOA

Drug Information (ALL DrugBank tools use query= as the search parameter, plus case_sensitive=False, exact_match=False, limit=N):

• drugbank_get_drug_basic_info_by_drug_name_or_id - Basic info. Param: query="drug_name"
• drugbank_get_indications_by_drug_name_or_drugbank_id - Approved indications. Param: query="drug_name"
• drugbank_get_pharmacology_by_drug_name_or_drugbank_id - Pharmacology. Param: query="drug_name"
• drugbank_get_targets_by_drug_name_or_drugbank_id - Drug targets. Param: query="drug_name"

Safety:

• FDA_get_warnings_and_cautions_by_drug_name - FDA warnings
• FAERS_search_reports_by_drug_and_reaction - Adverse events. Param: medicinalproduct= (NOT drug_name=)
• FAERS_count_death_related_by_drug - Serious outcomes. Param: medicinalproduct= (NOT drug_name=)
• drugbank_get_drug_interactions_by_drug_name_or_id - Interactions

Property Prediction:

• ADMETAI_predict_physicochemical_properties / ADMETAI_predict_toxicity - ADMET and toxicity

Pathway & Network Analysis:

• ReactomeAnalysis_pathway_enrichment - Pathway enrichment. Param: identifiers="SOD1\nTARDBP\nFUS" (newline-separated string, NOT array)
• STRING_get_network - Protein interaction networks. Param: identifiers="SOD1\rTARDBP\rFUS" (CR-separated string), species=9606
• CTD_get_gene_diseases - Curated gene-disease associations. Param: input_terms="gene_symbol" (NOT gene_symbol=)

Literature & Clinical Trials:

• PubMed_search_articles / EuropePMC_search_articles - Literature search
• search_clinical_trials - ClinicalTrials.gov search. Use condition for disease name. The intervention filter is strict and may miss trials — use query_term for broader drug-name matching as fallback.

CNS diseases note: For neurological indications (ALS, Alzheimer's, Parkinson's), prioritize BBB-penetrant candidates. Use ChEMBL molecular properties (MW < 500, PSA < 90) as BBB proxy since ADMETAI_predict_BBB_penetrance may require the tooluniverse[ml] extra. Consider route of administration (oral preferred for patients with swallowing difficulty) and sex-specific effects from preclinical models.

Scoring & Decision Framework

Repurposing Viability Score (0-100)

Classification:

• 80-100: Strong candidate — proceed to clinical evaluation
• 60-79: Promising — worth preclinical validation or retrospective study
• 40-59: Speculative — needs significant additional evidence
• <40: Weak — likely not worth pursuing without new mechanistic insight

Evidence Grading for Repurposing

How to Interpret and Combine Results

After running Phases 1-4, synthesize by answering:

1. Is the target validated for this disease? Check OpenTargets association score (>0.5 = strong). Cross-reference with genetic evidence (GWAS hits, rare variant studies). If target association is only pathway-level, the repurposing hypothesis is speculative.

2. Does the drug actually hit the target at achievable doses? Check ChEMBL IC50/Ki values. If the drug's affinity for the new target is >10x weaker than for its original target, clinical efficacy is unlikely at safe doses.

3. What's the safety margin? Compare the dose needed for the new indication to the approved dose. If higher doses are needed, safety data from the original indication may not apply.

4. Is there prior clinical evidence? A Phase II trial for the new indication (even failed) is more informative than 100 computational predictions. Check search_clinical_trials first.

5. What's the competitive landscape? If better drugs already exist for the disease, repurposing offers little value. Check DrugBank indications for approved therapies.

Best Practices

1. Check clinical trials FIRST: search_clinical_trials(condition="[disease]", intervention="[drug]") — if a trial already exists, start there
2. Validate targets with genetics: Genetic evidence (GWAS, rare variants) is the strongest predictor of successful drug development
3. Safety first: Prioritize approved drugs with known safety profiles
4. Dose matters: A drug that hits a disease target at 100x its approved dose is not a repurposing candidate
5. Mechanism over correlation: Network proximity alone is insufficient — explain WHY the drug should work
6. Consider IP and formulation: Generic drugs are easier to repurpose but harder to fund trials for

Computational Procedure: Drug-Target Dose Feasibility Check

A drug that hits a new target onl

Content truncated.