ToolUniverse

Overview

• Access 600+ scientific tools, models, datasets, and APIs
• Discover tools using natural language, semantic search, or keywords
• Execute tools through standardized AI-Tool Interaction Protocol
• Compose multi-step workflows for complex research problems
• Integration with Claude Desktop/Code via Model Context Protocol (MCP)

When to Use This Skill

• Searching for scientific tools by function or domain (e.g., "find protein structure prediction tools")
• Executing computational biology workflows (e.g., disease target identification, drug discovery, genomics analysis)
• Accessing scientific databases (OpenTargets, PubChem, UniProt, PDB, ChEMBL, KEGG, etc.)
• Composing multi-step research pipelines (e.g., target discovery → structure prediction → virtual screening)
• Working with bioinformatics, cheminformatics, or structural biology tasks
• Analyzing gene expression, protein sequences, molecular structures, or clinical data
• Performing literature searches, pathway enrichment, or variant annotation
• Building automated scientific research workflows

Quick Start

Basic Setup

from tooluniverse import ToolUniverse

# Initialize and load tools
tu = ToolUniverse()
tu.load_tools()  # Loads 600+ scientific tools

# Discover tools
tools = tu.run({
    "name": "Tool_Finder_Keyword",
    "arguments": {
        "description": "disease target associations",
        "limit": 10
    }
})

# Execute a tool
result = tu.run({
    "name": "OpenTargets_get_associated_targets_by_disease_efoId",
    "arguments": {"efoId": "EFO_0000537"}  # Hypertension
})

Model Context Protocol (MCP)

tooluniverse-smcp

Core Workflows

1. Tool Discovery

• Tool_Finder

  - Embedding-based semantic search (requires GPU)

• Tool_Finder_LLM

  - LLM-based semantic search (no GPU required)

• Tool_Finder_Keyword

  - Fast keyword search

# Search by natural language description
tools = tu.run({
    "name": "Tool_Finder_LLM",
    "arguments": {
        "description": "Find tools for RNA sequencing differential expression analysis",
        "limit": 10
    }
})

# Review available tools
for tool in tools:
    print(f"{tool['name']}: {tool['description']}")

references/tool-discovery.md

• Detailed discovery methods and search strategies
• Domain-specific keyword suggestions
• Best practices for finding tools

2. Tool Execution

# Execute disease-target lookup
targets = tu.run({
    "name": "OpenTargets_get_associated_targets_by_disease_efoId",
    "arguments": {"efoId": "EFO_0000616"}  # Breast cancer
})

# Get protein structure
structure = tu.run({
    "name": "AlphaFold_get_structure",
    "arguments": {"uniprot_id": "P12345"}
})

# Calculate molecular properties
properties = tu.run({
    "name": "RDKit_calculate_descriptors",
    "arguments": {"smiles": "CCO"}  # Ethanol
})

references/tool-execution.md

• Real-world execution examples across domains
• Tool parameter handling and validation
• Result processing and error handling
• Best practices for production use

3. Tool Composition and Workflows

# 1. Find disease targets
targets = tu.run({
    "name": "OpenTargets_get_associated_targets_by_disease_efoId",
    "arguments": {"efoId": "EFO_0000616"}
})

# 2. Get protein structures
structures = []
for target in targets[:5]:
    structure = tu.run({
        "name": "AlphaFold_get_structure",
        "arguments": {"uniprot_id": target['uniprot_id']}
    })
    structures.append(structure)

# 3. Screen compounds
hits = []
for structure in structures:
    compounds = tu.run({
        "name": "ZINC_virtual_screening",
        "arguments": {
            "structure": structure,
            "library": "lead-like",
            "top_n": 100
        }
    })
    hits.extend(compounds)

# 4. Evaluate drug-likeness
drug_candidates = []
for compound in hits:
    props = tu.run({
        "name": "RDKit_calculate_drug_properties",
        "arguments": {"smiles": compound['smiles']}
    })
    if props['lipinski_pass']:
        drug_candidates.append(compound)

references/tool-composition.md

• Complete workflow examples (drug discovery, genomics, clinical)
• Sequential and parallel tool composition patterns
• Output processing hooks
• Workflow best practices

Scientific Domains

• Sequence analysis, alignment, BLAST
• Gene expression (RNA-seq, DESeq2)
• Pathway enrichment (KEGG, Reactome, GO)
• Variant annotation (VEP, ClinVar)

• Molecular descriptors and fingerprints
• Drug discovery and virtual screening
• ADMET prediction and drug-likeness
• Chemical databases (PubChem, ChEMBL, ZINC)

• Protein structure prediction (AlphaFold)
• Structure retrieval (PDB)
• Binding site detection
• Protein-protein interactions

• Mass spectrometry analysis
• Protein databases (UniProt, STRING)
• Post-translational modifications

• Genome assembly and annotation
• Copy number variation
• Clinical genomics workflows

• Disease databases (OpenTargets, OMIM)
• Clinical trials and FDA data
• Variant classification

references/domains.md

• Complete domain categorization
• Tool examples by discipline
• Cross-domain applications
• Search strategies by domain

Reference Documentation

• references/installation.md

  - Installation, setup, MCP configuration, platform integration

• references/tool-discovery.md

  - Discovery methods, search strategies, listing tools

• references/tool-execution.md

  - Execution patterns, real-world examples, error handling

• references/tool-composition.md

  - Workflow composition, complex pipelines, parallel execution

• references/domains.md

  - Tool categorization by domain, use case examples

• references/api_reference.md

  - Python API documentation, hooks, protocols

references/tool-discovery.md

Example Scripts

scripts/example_tool_search.py

• Keyword-based search
• LLM-based search
• Domain-specific searches
• Getting detailed tool information

scripts/example_workflow.py

• Drug discovery pipeline (disease → targets → structures → screening → candidates)
• Genomics analysis (expression data → differential analysis → pathways)

Best Practices

1. Tool Discovery:
   • Start with broad searches, then refine based on results
   • Use

     Tool_Finder_Keyword

     for fast searches with known terms
   • Use

     Tool_Finder_LLM

     for complex semantic queries
   • Set appropriate

     limit

     parameter (default: 10)
2. Tool Execution:
   • Always verify tool parameters before execution
   • Implement error handling for production workflows
   • Validate input data formats (SMILES, UniProt IDs, gene symbols)
   • Check result types and structures
3. Workflow Composition:
   • Test each step individually before composing full workflows
   • Implement checkpointing for long workflows
   • Consider rate limits for remote APIs
   • Use parallel execution when tools are independent
4. Integration:
   • Initialize ToolUniverse once and reuse the instance
   • Call

     load_tools()

     once at startup
   • Cache frequently used tool information
   • Enable logging for debugging

Key Terminology

• Tool: A scientific resource (model, dataset, API, package) accessible through ToolUniverse
• Tool Discovery: Finding relevant tools using search methods (Finder, LLM, Keyword)
• Tool Execution: Running a tool with specific arguments via

  tu.run()

• Tool Composition: Chaining multiple tools for multi-step workflows
• MCP: Model Context Protocol for integration with Claude Desktop/Code
• AI-Tool Interaction Protocol: Standardized interface for LLM-tool communication

Resources

• Official Website: https://aiscientist.tools
• GitHub: https://github.com/mims-harvard/ToolUniverse
• Documentation: https://zitniklab.hms.harvard.edu/ToolUniverse/
• Installation:

  uv uv pip install tooluniverse

• MCP Server:

  tooluniverse-smcp