Evidence Methodology — How WePattern Sources Health Intelligence
WePattern maps 3,500+ clinically documented health relationships. Every relationship in the knowledge graph requires at least one published clinical source before it is included.
Primary Sources
- PubMed-indexed clinical research — peer-reviewed studies published in indexed medical journals
- NIH Office of Dietary Supplements — authoritative nutrient fact sheets and clinical guidance
- Natural Medicines Database — evidence-graded drug-nutrient interaction data
- Functional medicine clinical references — Institute for Functional Medicine, peer-reviewed functional medicine literature
Evidence Grading
Relationships in the knowledge graph are graded by evidence quality:
- Grade A — Multiple randomized controlled trials or meta-analyses confirming the relationship
- Grade B — At least one high-quality observational study or controlled trial supporting the relationship
- Grade C — Clinical case series, expert consensus, or mechanistic evidence with clinical plausibility
WePattern surfaces Grade A and B evidence as primary signals. Grade C is included for completeness and flagged accordingly.
What WePattern Does Not Do
WePattern does not generate answers from a general language model without grounding in mapped data. Every surfaced relationship traces back to a specific clinical source. WePattern does not diagnose, prescribe, or recommend — it surfaces patterns and educational context for discussion with a qualified healthcare provider.
Knowledge Graph Maintenance
The knowledge graph is updated continuously as new research is published. Relationships are reviewed for currency and accuracy. When new evidence updates or contradicts an existing mapped relationship, the knowledge graph is revised.