Available Tools Reference

Complete reference of all ToolUniverse scientific tools and their capabilities.

ToolUniverse provides 1000+ tools across eight major categories, each serving specific computational and analytical requirements in scientific research.

Tool Ecosystem Overview

ToolUniverse integrates tools across eight major categories:

ToolUniverse Ecosystem (1000+ Tools):

┌─────────────────┐
│   ML Models     │ 15 tools  → Prediction, Classification, Generation
│     (AI/ML)     │
└─────────────────┘

┌─────────────────┐
│   AI Agents     │ 33 tools  → Autonomous Planning, Tool Routing
│   (Agentic)     │
└─────────────────┘

┌─────────────────┐
│   Software      │ 164 tools → Bioinformatics, Analysis Packages
│   Packages      │
└─────────────────┘

┌─────────────────┐
│ Human Expert    │ 6 tools   → Consultation, Validation, Feedback
│   Feedback      │
└─────────────────┘

┌─────────────────┐
│   Robotics      │ 1 tool    → ROS Communication, Lab Automation
│  (Automation)   │
└─────────────────┘

┌─────────────────┐
│   Databases     │ 84 tools  → Structured Data, Knowledge Bases
│   (Storage)     │
└─────────────────┘

┌─────────────────┐
│  Embedding      │ 4 tools   → Vector Search, Semantic Retrieval
│   Stores        │
└─────────────────┘

┌─────────────────┐
│      APIs       │ 281 tools → External Services, Data Access
│  (Integration)  │
└─────────────────┘

Tool Categories Summary

Tool Distribution by Category

Category	Count	Percentage	Primary Use Cases
APIs	281	48.4%	External data access, real-time information
Software Packages	164	28.3%	Computational analysis, local processing
Databases	84	14.5%	Structured data storage and retrieval
AI Agents	33	5.7%	Autonomous reasoning and planning
ML Models	15	2.6%	Prediction and classification tasks
Expert Feedback	6	1.0%	Human validation and guidance
Embedding Stores	4	0.7%	Semantic search and similarity
Robotics	1	0.2%	Laboratory automation
Total	588	100%	Comprehensive scientific ecosystem

Molecular & Genetic Data

UniProt - Protein Information

Access comprehensive protein and gene information.

Key Functions: * UniProt_get_function_by_accession - Get functional annotations by UniProt accession * UniProt_search - Search UniProtKB with field queries * UniProt_get_sequence_by_accession - Retrieve protein sequences

Example:

query = {
    "name": "UniProt_get_function_by_accession",
    "arguments": {"accession": "P38398"}  # BRCA1 accession
}
result = tu.run(query)

Gene Ontology - Functional Annotation

Gene Ontology annotations and functional analysis.

Key Functions: * GO_get_annotations_for_gene - Get GO annotations for a gene * GO_search_terms - Search GO terms * GO_get_genes_for_term - Get genes annotated to a GO term

Example:

query = {
    "name": "GO_get_annotations_for_gene",
    "arguments": {"gene_id": "TP53"}
}

Enrichr - Gene Set Analysis

Comprehensive gene set enrichment analysis.

Key Functions: * Enrichr_enrich - Enrichment analysis for a gene list against one library * Enrichr_list_libraries - List available gene set libraries * Enrichr_get_top_enriched - Top enriched terms across several libraries

Example:

query = {
    "name": "Enrichr_enrich",
    "arguments": {
        "gene_list": ["BRCA1", "BRCA2", "TP53", "ATM", "CHEK2"],
        "library": "KEGG_2021_Human"
    }
}

Disease & Target Data

OpenTargets Platform

Comprehensive disease-target association data.

Key Functions: * OpenTargets_get_associated_targets_by_disease_efoId - Disease-associated targets * OpenTargets_get_associated_diseases_by_drug_chemblId - Drug-associated diseases * OpenTargets_get_disease_id_description_by_name - Disease lookup * OpenTargets_get_evidence_by_datasource - Evidence for associations * OpenTargets_get_drug_mechanisms_of_action_by_chemblId - Drug mechanisms of action

Example:

# Get targets for Alzheimer's disease
query = {
    "name": "OpenTargets_get_associated_targets_by_disease_efoId",
    "arguments": {"efoId": "MONDO_0004975"}  # Alzheimer disease
}

EFO - Experimental Factor Ontology

Disease and experimental factor ontology.

Key Functions: * ols_search_efo_terms - Search EFO terms by name * OSL_get_efo_id_by_disease_name - Resolve a disease name to an EFO ID * ols_get_efo_term_children - Get child terms of an EFO term

Example:

query = {
    "name": "ols_search_efo_terms",
    "arguments": {"query": "diabetes mellitus", "rows": 5}
}

Drug & Chemical Data

PubChem - Chemical Information

Comprehensive chemical compound database.

Key Functions: * PubChem_get_CID_by_compound_name - Look up compound IDs by name * PubChem_search_compounds_by_substructure - Search compounds by substructure * PubChem_get_compound_properties_by_CID - Molecular properties * PubChem_search_compounds_by_similarity - Chemical similarity search

Example:

query = {
    "name": "PubChem_get_CID_by_compound_name",
    "arguments": {"name": "aspirin"}
}

ChEMBL - Bioactivity Data

Chemical bioactivity and drug discovery data.

Key Functions: * ChEMBL_get_molecule_targets - Get targets for a molecule * ChEMBL_search_targets - Find target ChEMBL IDs by name or gene symbol * ChEMBL_get_target_activities - Bioactivity measurements for a target * ChEMBL_search_similar_molecules - Chemical similarity search

Example:

# EGFR is CHEMBL203; use ChEMBL_search_targets to look an ID up by name
query = {
    "name": "ChEMBL_get_target_activities",
    "arguments": {"target_chembl_id": "CHEMBL203", "limit": 20}
}

️ Drug Safety & Regulatory

OpenFDA - FDA Data

FDA drug labeling and adverse event data.

Key Functions: * FAERS_count_reactions_by_drug_event - Count adverse reactions by drug * FDA_get_warnings_by_drug_name - Get FDA warnings * OpenFDA_search_drug_labels - Drug labeling information * OpenFDA_search_drug_enforcement - Drug recall and enforcement information

Example:

# Search adverse events
query = {
    "name": "FAERS_count_reactions_by_drug_event",
    "arguments": {"medicinalproduct": "warfarin"}
}

# Get FDA warnings
query = {
    "name": "FDA_get_warnings_by_drug_name",
    "arguments": {"drug_name": "warfarin"}
}

DailyMed - Drug Labeling

Official FDA drug labeling information.

Key Functions: * DailyMed_search_spls - Search structured product labels by drug name, NDC or RxCUI * DailyMed_get_spl_by_setid - Get a full label by its set ID * DailyMed_parse_dosing - Extract the dosing section from a label

Example:

query = {
    "name": "DailyMed_search_spls",
    "arguments": {"drug_name": "metformin"}
}

Clinical Research

ClinicalTrials.gov

Clinical trial registry and results database.

Key Functions: * ClinicalTrials_search_studies - Search clinical trials * ClinicalTrials_get_study - Get detailed study information by NCT ID * ClinicalTrials_search_by_intervention - Find trials by intervention * ClinicalTrials_search_by_sponsor - Find trials by lead sponsor

Example:

query = {
    "name": "ClinicalTrials_search_studies",
    "arguments": {
        "query_cond": "breast cancer",
        "query_intr": "immunotherapy"
    }
}

Literature & Publications

PubTator - Biomedical Literature

PubMed literature with named entity recognition.

Key Functions: * PubTator3_LiteratureSearch - Search literature with entities * PubTator3_get_annotations - Get entity annotations for PMIDs * PubTator3_EntityAutocomplete - Resolve a free-text name to a PubTator entity ID

Example:

query = {
    "name": "PubTator3_LiteratureSearch",
    "arguments": {
        "query": "@GENE_BRCA1 AND @DISEASE_Neoplasms"
    }
}

Europe PMC

European literature database with full-text access.

Key Functions: * EuropePMC_search_articles - Search articles and abstracts * EuropePMC_get_full_text - Get full-text when available * EuropePMC_get_citations - Get citation data

Example:

query = {
    "name": "EuropePMC_search_articles",
    "arguments": {"query": "CRISPR gene therapy"}
}

Semantic Scholar

AI-powered academic search engine.

Key Functions: * SemanticScholar_search_papers - Search academic papers * SemanticScholar_get_paper - Get detailed paper information * SemanticScholar_get_paper_citations - Citation network analysis

Example:

query = {
    "name": "SemanticScholar_search_papers",
    "arguments": {"query": "machine learning drug discovery"}
}

OpenAlex

Open academic publication database.

Key Functions: * openalex_search_works - Search academic works * openalex_get_author - Author information and metrics * openalex_get_institution - Institution research data

Specialized Databases

Human Protein Atlas

Tissue and cell expression data.

Key Functions: * HPA_get_rna_expression_in_specific_tissues - Tissue expression patterns * HPA_get_comparative_expression_by_gene_and_cellline - Cell-line expression data * HPA_get_subcellular_location - Subcellular localization

Example:

query = {
    "name": "HPA_get_rna_expression_in_specific_tissues",
    "arguments": {
        "ensembl_id": "ENSG00000012048",  # BRCA1
        "tissue_names": ["breast", "ovary"]
    }
}

Reactome Pathways

Biological pathway database.

Key Functions: * Reactome_map_uniprot_to_pathways - Pathways for a protein * ReactomeContent_search - Search pathway database * Reactome_get_pathway - Detailed pathway information

Example:

query = {
    "name": "Reactome_map_uniprot_to_pathways",
    "arguments": {"uniprot_id": "P04637"}  # TP53
}

HumanBase

Tissue-specific gene networks.

Key Functions: * humanbase_ppi_analysis - Tissue-specific protein-protein interaction networks

MedlinePlus

Consumer health information.

Key Functions: * MedlinePlus_search_topics_by_keyword - Health topic information * MedlinePlus_get_genetics_condition_by_name - Genetic condition information * MedlinePlus_connect_lookup_by_code - Look up consumer information by drug or test code

AI-Powered Tools

Machine Learning Models (15 tools)

Apply machine learning algorithms for prediction, classification, and generation tasks.

Core ML Tools:

boltz2_docking - Protein-ligand binding prediction

{
    "name": "boltz2_docking",
    "arguments": {
        "sequence": "MVLSPADKTNVKAAW",
        "ligands": ["CCO"],
        "recycling_steps": 3,
        "sampling_steps": 200,
        "diffusion_samples": 1,
        "step_scale": 1.638,
        "use_potentials": False,
        "return_structure": True
    }
}
# Returns: binding affinity and probability, plus the predicted structure

ADMETAI_predict_CYP_interactions - Drug metabolism prediction

{
    "name": "ADMETAI_predict_CYP_interactions",
    "arguments": {
        "smiles": ["CC(=O)OC1=CC=CC=C1C(=O)O"]  # Aspirin
    }
}
# Returns: per-CYP-isoform inhibition probabilities

AI Agents (33 tools)

Autonomous tools that perceive environments, make decisions, and take actions toward research goals.

Literature & Analysis Agents:

HypothesisGenerator - Generate research hypotheses

{
    "name": "HypothesisGenerator",
    "arguments": {
        "context": "Checkpoint inhibitors work in only a subset of solid tumours.",
        "domain": "cancer immunotherapy",
        "number_of_hypotheses": 5
    }
}
# Returns: ranked hypotheses with supporting rationale

ExperimentalDesignScorer - Evaluate experimental designs

{
    "name": "ExperimentalDesignScorer",
    "arguments": {
        "hypothesis": "EGFR inhibition slows progression in EGFR-mutant NSCLC.",
        "design_description": "Phase II single-arm trial, 80 patients, 12-month PFS endpoint"
    }
}
# Returns: design score with improvement suggestions

MedicalLiteratureReviewer - Comprehensive literature analysis

{
    "name": "MedicalLiteratureReviewer",
    "arguments": {
        "research_topic": "CAR-T cell therapy safety profile",
        "literature_content": "<abstracts or full text to synthesise>",
        "focus_area": "safety profile",
        "study_types": "randomized controlled trials",
        "quality_level": "moderate and above",
        "review_scope": "rapid review"
    }
}
# Returns: synthesised review with key findings and research gaps

Search & Integration Tools

Tool Finder

Find appropriate tools for your research needs.

Key Functions: * Tool_Finder_Keyword - Keyword-based tool search * Tool_Finder_LLM - LLM-reasoned tool search * Tool_RAG - Embedding-based semantic tool search

Example:

query = {
    "name": "Tool_Finder_Keyword",
    "arguments": {"description": "drug safety and adverse events", "limit": 10}
}

Embedding Stores

Store and retrieve vectorized representations of scientific data for semantic search.

Core Embedding Tools:

embedding_database_create - Create a collection to embed into

{
    "name": "embedding_database_create",
    "arguments": {"database_name": "pubmed_abstracts"}
}

embedding_database_search - Vector similarity search

{
    "name": "embedding_database_search",
    "arguments": {
        "database_name": "pubmed_abstracts",
        "query": "protein folding dynamics",
        "top_k": 50
    }
}
# Returns: similar documents with relevance scores and metadata

️ Tool Usage Patterns

Single Tool Queries

Simple, focused queries for specific information:

# Get protein function by accession (EGFR → P00533)
protein_query = {
    "name": "UniProt_get_function_by_accession",
    "arguments": {"accession": "P00533"}
}

# Search adverse events
safety_query = {
    "name": "FAERS_count_reactions_by_drug_event",
    "arguments": {"medicinalproduct": "metformin"}
}

Multi-Tool Workflows

Combine multiple tools for comprehensive analysis:

# Step 1: Get disease info
disease_query = {
    "name": "OpenTargets_get_disease_id_description_by_name",
    "arguments": {"diseaseName": "diabetes"}
}

# Step 2: Get associated targets
targets_query = {
    "name": "OpenTargets_get_associated_targets_by_disease_efoId",
    "arguments": {"efoId": disease_id}
}

# Step 3: Analyze target pathways
pathway_query = {
    "name": "Enrichr_enrich",
    "arguments": {
        "gene_list": target_list,
        "library": "KEGG_2021_Human"
    }
}

Batch Processing

Process multiple related queries efficiently:

# Process multiple genes
genes = ["BRCA1", "BRCA2", "TP53", "ATM"]

results = {}
for accession in ["P38398", "P51587", "P04637", "Q13315"]:  # BRCA1, BRCA2, TP53, ATM
    query = {
        "name": "UniProt_get_function_by_accession",
        "arguments": {"accession": accession}
    }
    results[accession] = tu.run(query)

Integration Patterns

Multi-Tool Workflows

Combine multiple tools for comprehensive analysis:

from tooluniverse import ToolUniverse

# Drug discovery workflow
def drug_discovery_pipeline(disease_name):
    tooluni = ToolUniverse()
    tooluni.load_tools()

    # 1. Find disease ID
    disease_query = {
        "name": "OpenTargets_get_disease_id_description_by_name",
        "arguments": {"diseaseName": disease_name}
    }
    disease_info = tooluni.run(disease_query)

    # 2. Get associated targets
    targets_query = {
        "name": "OpenTargets_get_associated_targets_by_disease_efoId",
        "arguments": {"efoId": disease_info['id']}
    }
    targets_result = tooluni.run(targets_query)
    targets = targets_result['data']['disease']['associatedTargets']['rows']

    # 3. Find drugs for each target
    drugs = []
    for row in targets[:5]:  # Top 5 targets
        target = row['target']
        drugs_query = {
            "name": "OpenTargets_get_associated_drugs_by_target_ensemblID",
            "arguments": {"ensemblId": target['id']}
        }
        target_drugs = tooluni.run(drugs_query)
        drugs.extend(target_drugs)

    # 4. Check safety profiles
    for drug in drugs[:10]:  # Top 10 drugs
        safety_query = {
            "name": "FDA_get_warnings_by_drug_name",
            "arguments": {"drug_name": drug['name']}
        }
        safety = tooluni.run(safety_query)
        drug['safety_warnings'] = safety

    return drugs

Tool Composition Patterns

Sequential Workflows:

# Disease → Targets → Compounds → Prediction
workflow = [
    ("OpenTargets_get_associated_targets_by_disease_efoId", {"efoId": disease_id}),
    ("ChEMBL_get_target_activities", {"target_chembl_id": target_chembl_id}),
    ("ADMETAI_predict_physicochemical_properties", {"smiles": [compound]}),
    ("ADMETAI_predict_toxicity", {"smiles": [compound]})
]

Parallel Data Gathering:

# Multi-database literature search
parallel_searches = [
    ("PubTator3_LiteratureSearch", {"query": research_topic}),
    ("EuropePMC_search_articles", {"query": research_topic}),
    ("SemanticScholar_search_papers", {"query": research_topic})
]

Feedback Loops:

# Iterative optimization
while not satisfactory_result:
    prediction = ml_model_prediction(current_compound)
    if prediction.score < threshold:
        analogs = chemical_database_search(current_compound)
        current_compound = select_best_analog(analogs)
    else:
        break

Tool Performance Tips

Optimization Strategies

1. Use specific queries: More specific queries return faster

2. Limit results: Use limit parameter to control result size

3. Cache results: Enable caching for repeated queries

4. Batch when possible: Some tools support batch operations

Rate Limiting

ToolUniverse automatically handles API rate limits, but you can optimize:

import time

# Add delays for large batch operations
for query in large_query_list:
    result = tu.run(query)
    time.sleep(0.1)  # Small delay between requests

Error Handling

Always include error handling for robust applications:

try:
    result = tu.run(query)
    if result and 'data' in result:
        # Process successful result
        process_data(result['data'])
    else:
        print("No data returned")
except Exception as e:
    print(f"Query failed: {e}")

Performance Optimization

Category-Specific Considerations

ML Models: - Remote execution reduces local resource requirements - Batch predictions when possible - Cache results for expensive computations

APIs: - Respect rate limits and implement backoff - Use pagination for large datasets - Cache frequent queries

Databases: - Use specific field queries instead of full searches - Implement result limits for exploration - Index frequently accessed data

Agents: - Configure appropriate timeout values - Use streaming for long-running tasks - Implement progress monitoring

Best Practices

1. Tool Selection: Choose the right tool for your specific use case

2. Rate Limiting: Respect API rate limits to avoid blocking

3. Error Handling: Always handle potential API errors gracefully

4. Caching: Use caching for frequently accessed data

5. Batch Processing: Use batch operations when available for efficiency

6. Configuration: Configure tools appropriately for your environment

Tool Discovery & Selection

Finding the Right Tools

By Category:

# List tools by type (use get_tool_types() to see available types)
print(tu.get_tool_types())        # e.g. ['opentarget', 'ChEMBL', 'uniprot', ...]
ml_tools = tu.filter_tools(include_tool_types=["admetai"])
database_tools = tu.filter_tools(include_tool_types=["uniprot", "ChEMBL"])
api_tools = tu.filter_tools(include_tool_types=["EuropePMC", "pubtator"])

By Functionality:

# Keyword search across all categories
protein_tools = tu.run({
    "name": "Tool_Finder_Keyword",
    "arguments": {"description": "protein structure prediction", "limit": 10}
})
drug_tools = tu.run({
    "name": "Tool_Finder_Keyword",
    "arguments": {"description": "drug safety analysis", "limit": 10}
})
literature_tools = tu.run({
    "name": "Tool_Finder_Keyword",
    "arguments": {"description": "literature review automation", "limit": 10}
})

By Domain:

# Load domain-specific tools
tu.load_tools(tool_type=[
    "opentarget",    # Disease-target data
    "ChEMBL",        # Chemical data
    "uniprot",       # Protein data
    "pubtator"       # Literature with entities
])

API Authentication

# API keys are managed via environment variables
# Set them before importing ToolUniverse or use a .env file
import os

os.environ['NCBI_API_KEY'] = 'your_ncbi_key'
os.environ['SEMANTIC_SCHOLAR_API_KEY'] = 'your_s2_key'

# ToolUniverse automatically reads API keys from environment variables
tu = ToolUniverse()
tu.load_tools()

Future Extensions

Planned Categories: - Visualization Tools: Interactive plotting and dashboard generation - Workflow Engines: Advanced orchestration and scheduling - Cloud Services: Distributed computing and storage - Compliance Tools: Regulatory and ethics validation

Community Contributions: - Tool submission guidelines - Quality assurance processes - Community voting and validation - Maintenance and updates

Next Steps

Now that you know what tools are available:

• Try Examples: Examples - See tools in action

• Build Workflows: Scientific Workflows - Combine tools for research

• Extend ToolUniverse: Navigation - Create custom tools

Tip

Discovery tip: Use the AI-powered tool discovery features to find the right tools for your specific research questions!

Tip

Tool ecosystem synergy: The eight categories are designed to work together. APIs provide data access, ML models add intelligence, agents orchestrate complex workflows, while databases and embedding stores enable efficient information management.