An Ontology‑Guided Drug–Herb–Food Interaction Checker with Mechanism‑Based Knowledge Graph Reasoning and Condition‑Aware Interpretation

Drug-centric interaction checkers: Existing interaction-checking systems are largely focused on drugs and do not adequately cover drug–herb–food co-consumption patterns.

Limited mechanistic transparency: Existing systems rely on predefined interaction pairs and provide limited mechanistic explanations.

Limited condition-aware interpretation: Existing interaction checkers are poorly equipped to represent interactions influenced by health-related conditions.

Drug-centric interaction checkers: Existing interaction-checking systems are largely focused on drugs and do not adequately cover drug–herb–food co-consumption patterns.

Limited mechanistic transparency: Existing systems rely on predefined interaction pairs and provide limited mechanistic explanations.

Ontology-guided interaction knowledge graph: The authors developed DHFI-C as an ontology-guided knowledge graph platform for drug–herb–food interaction assessment.

Mechanism-based deterministic inference: The platform uses a graph-native representation with a deterministic inference engine to derive pharmacokinetic and pharmacodynamic interactions through shared mechanisms.

Condition-inclusive biomedical data model: The system models drugs, herbs, foods, health-related conditions, and diseases in a structured ontology-aligned data model.

Curated evidence-based interaction resource: Evidence from open-access literature was curated under PRISMA 2020 and transformed into the system data model.

Ontology-guided interaction knowledge graph: The authors developed DHFI-C as an ontology-guided knowledge graph platform for drug–herb–food interaction assessment.

Large integrated interaction graph:

Provenance-backed inferred interaction reporting:

Use-case condition-aware performance: