KB‑HD provides full REST API access for programmatic exploration.

• Key Endpoints:
  • /entities – search by name, ID, filters
  • /relationships – query specific entity pairs or types
• Supports: JSON and CSV output, pagination, filters (type, species)
• Security: Use your unique API key (in header); rate limits apply
• Sample Code (Python):
  • python
CopyEdit
import requests
• headers = {"Authorization": "Bearer YOUR_API_KEY"}
• params = {"query": "BRCA1", "type": "gene"}
• r = requests.get("https://api.qiagen.com/entities", headers=headers, params=params)

• Learn More: See our “Quick-start code snippets” section (Python, R, cURL) at MyQDI (Access needed)

Even though BKB‑AI is powered by knowledge graphs with AI-derived edges, you can still interrogate metadata via standard SQL tables. Graph‑specific querying (e.g., via specialized APIs or graph databases) is available when licensed, but relational querying is a great first step for schema discovery and basic filtering.

Performance Tips & Indexing 

• Default imports include indexes optimized for standard lookups
• For heavy queries on aggregated relationships (e.g., molecule ↔ disease/function), you can optionally create:

sql

CopyEdit

CREATE INDEX idx_aggregated_molecule_disease_and_function_id
ON molecule_disease_and_function_relationships_aggregated_ids(aggregated_relationship_id);

• Adds ~7 GB to DB size; use only if needed for performance

Best Practices & Caution

• Don’t modify the core SQLite build; altering tables or schema may break packaged tools.
• Use pivot views only for display, not for filtering.
• Explore system tables first when navigating new datasets.

Quick Start Summary 

• Unzip the SQLite build.
• Open in your SQL tool or connect via R/Python wrappers.
• Discover schema via tables & table_columns.
• Explore sample data via metadata tables.
• Display via pivot views.
• Optional: add indexing for speed.
• For richer graph or AI-driven relationships, consider moving into graph-based tools when licensed.