network-pharmacology

Files (3)

• SKILL.md
• run.cpython-313.pyc · scripts/__pycache__/
• run.py · scripts/

SKILL.md (first lines)

---
name: tooluniverse-network-pharmacology
description: Construct and analyze compound-target-disease networks for drug repurposing, polypharmacology discovery, and systems pharmacology. Builds multi-layer networks from ChEMBL, OpenTargets, STRING, DrugBank, Reactome, FAERS, and 60+ other ToolUniverse tools. Calculates Network Pharmacology Scores (0-100), identifies repurposing candidates, predicts mechanisms, and analyzes polypharmacology. Use when users ask about drug repurposing via network analysis, multi-target drug effects, compound-target-disease networks, systems pharmacology, or polypharmacology.
---

# Network Pharmacology Pipeline

Construct and analyze compound-target-disease (C-T-D) networks to identify drug repurposing opportunities, understand polypharmacology, and predict drug mechanisms using systems pharmacology approaches.

**IMPORTANT**: Always use English terms in tool calls (drug names, disease names, target names), even if the user writes in another language. Only try original-language terms as a fallback if English returns no results. Respond in the user's language.

---

## When to Use This Skill

Apply when users:
- Ask "Can [drug] be repurposed for [disease] based on network analysis?"
- Want to understand multi-target (polypharmacology) effects of a compound
- Need compound-target-disease network construction and analysis
- Ask about network proximity between drug targets and disease genes
- Want systems pharmacology analysis of a drug or target
- Ask about drug repurposing candidates ranked by network metrics
- Need mechanism prediction for a drug in a new indication
- Want to identify hub genes in disease networks as therapeutic targets
- Ask about disease module coverage by a compound's targets

**NOT for** (use other skills instead):
- Simple drug repurposing without network analysis -> Use `tooluniverse-drug-repurposing`
- Single target validation -> Use `tooluniverse-drug-target-validation`
- Adverse event detection only -> Use `tooluniverse-adverse-event-detection`
- General disease research -> Use `tooluniverse-disease-research`
- GWAS interpretation -> Use `tooluniverse-gwas-snp-interpretation`

---

## Input Parameters

| Parameter | Required | Description | Example |
|-----------|----------|-------------|---------|
| **entity** | Yes | Compound name/ID, target gene symbol/ID, or disease name/ID | `metformin`, `EGFR`, `Alzheimer disease` |
| **entity_type** | No | Type hint: `compound`, `target`, or `disease` (auto-detected if omitted) | `compound` |
| **analysis_mode** | No | `compound-to-disease`, `disease-to-compound`, `target-centric`, `bidirectional` (default) | `bidirectional` |
| **secondary_entity** | No | Second entity for focused analysis (e.g., disease for compound input) | `Alzheimer disease` |

---

## Network Pharmacology Score (0-100)

### Score Components

**Network Proximity** (0-35 points):
- Strong proximity (Z < -2, p < 0.01): 35 points
- Moderate proximity (Z < -1, p < 0.05): 20 points
- Weak proximity (Z < -0.5): 10 points
- No proximity: 0 points

**Clinical Evidence** (0-25 points):
- Approved for related indication: 25 points
- Active clinical trials: 15 points
- Completed trials with positive results: 10 points
- Preclinical only: 5 points

**Target-Disease Association** (0-20 points):
- Strong genetic evidence (GWAS, rare variants): 20 points
- Moderate evidence (pathways, literature): 12 points
- Weak evidence (computational only): 5 points

**Safety Profile** (0-10 points):
- FDA-approved, favorable safety: 10 points
- Known manageable adverse events: 7 points
- Significant safety concerns: 3 points
- Black box warning relevant to indication: 0 points

**Mechanism Plausibility** (0-10 points):
- Clear pathway mechanism with functional evidence: 10 points
- Indirect mechanism via network neighbors: 6 points
- Purely computational prediction: 2 points

### Priority Tiers

| Score | Tier | Recommendation |
|-------|------|----------------|
| **80-100** | Tier 1 | High repurposing potential - proceed with experimental validation |
| **60-79** | Tier 2 | Good potential - needs mechanistic validation |
| **40-59** | Tier 3 | Moderate potential - high-risk/high-reward, needs extensive validation |
| **0-39** | Tier 4 | Low potential - consider alternative approaches |

### Evidence Grading System

| Tier | Symbol | Criteria | Examples |
|------|--------|----------|----------|
| **T1** | [T1] | Human clinical proof, regulatory evidence | FDA-approved indication, Phase III trial, patient genomics |
| **T2** | [T2] | Functional experimental evidence | Bioactivity data (IC50 < 1 uM), CRISPR screen, animal model |
| **T3** | [T3] | Association/computational evidence | GWAS hit, network proximity, pathway enrichment, expression |
| **T4** | [T4] | Prediction, annotation, text-mining | AlphaFold prediction, database annotation, literature co-mention |

---

## KEY PRINCIPLES

1. **Report-first approach** - Create report file FIRST, then populate progressively
2. **Entity disambiguation FIRST** - Resolve all identifiers before analysis
3. **Bidirectional network** - Construct C-T-D network comprehensively from both directions
4. **Network metrics** - Calculate proximity, centrality, module overlap quantitatively
5. **Rank candidates** - Prioritize by composite Network Pharmacology Score
6. **Mechanism prediction** - Explain HOW drug could work for disease via network paths
7. **Clinical feasibility** - FDA-approved drugs ranked higher than preclinical
8. **Safety context** - Flag known adverse events and off-target liabilities
9. **Evidence grading** - Grade all evidence T1-T4
10. **Negative results documented** - "No data" is data; empty sections are failures
11. **Source references** - Every finding must cite the source tool/database
12. **Completeness checklist** - Mandatory section at end showing analysis coverage

---

## Complete Workflow

### Phase 0: Entity Disambiguation and Report Setup

From lamm-mit/scienceclaw, Apache-2.0.

More from lamm-mit/scienceclaw