BioKG-Builder is an AI-driven biomedical literature knowledge graph generator that helps researchers analyze and visualize relationships between biological entities from PubMed literature.

• Automated Literature Search: Search PubMed database for relevant articles
• AI-Powered Analysis: Use LLM to identify causal relationships between biological entities
• Entity Recognition: Extract and deduplicate biological entities (genes, proteins, diseases, etc.)
• Knowledge Graph Visualization: Create interactive network visualizations using Pyvis
• Intelligent Filtering: Focus on specific entities of interest
• Automated Reporting: Generate comprehensive analysis reports using AI

git clone https://github.com/Zaoqu-Liu/biokg-builder.git
cd biokg-builder
pip install -e .

from biokg_builder import BioKGBuilder

# Initialize BioKGBuilder
builder = BioKGBuilder(
    email="your.email@example.com",        # Replace with your email
    api_key="your-api-key",         # Replace with your LLM API Key
    base_url="https://api.deepseek.com",   # Optional: custom API endpoint
    use_parallel=True                      # Enable parallel processing
)

# Build a knowledge graph for a specific keyword
results = builder.build_knowledge_graph("THBS2")

# View results
print(f"Construction completed with {len(results['entities'])} biological entities found.")
print(f"Generated files: {results['files']}")

The tool generates several files:

1. Excel Files:

   • {keyword}_pubmed_search_results.xlsx: Raw PubMed search results
   • modified_updated_{keyword}_causal.xlsx: Processed causal relationships

2. HTML Visualizations:

   • {keyword}_entity_network.html: Complete knowledge graph
   • filtered_entity_{keyword}_network.html: Filtered subgraph

3. Reports: AI-generated summary of findings

This project is licensed under the MIT License - see the LICENSE file for details.

If you use BioKG-Builder in your research, please cite:

@software{biokg_builder,
  title={BioKG-Builder: AI-driven biomedical literature knowledge graph generator},
  author={Zaoqu Liu},
  year={2025},
  url={https://github.com/Zaoqu-Liu/biokg-builder}
}

• BioPython team for PubMed access tools
• OpenAI for language models
• Pyvis developers for network visualization

• Author: Zaoqu Liu
• Email: liuzaoqu@163.com
• GitHub: @Zaoqu-Liu
• Issues: GitHub Issues