A Rule-Based Inference Framework to Explore and Explain the Biological Related Mechanisms of Potential Drug-Drug...

Area of Science:

• Pharmacology
• Computational Biology
• Biomedical Informatics

Background:

• Increasing polypharmacy necessitates better prediction of drug-drug interactions (DDIs).
• Traditional DDI identification methods struggle to elucidate interaction mechanisms and clinical relevance.
• Existing computational approaches have limited clinical utility.

Purpose of the Study:

• To develop a rule-based inference framework for predicting DDI mechanisms.
• To support the determination of the clinical relevance of potential DDIs.
• To integrate diverse biomedical resources for comprehensive DDI analysis.

Main Methods:

• Developed a rule-based inference framework utilizing a knowledge graph.
• Integrated biomedical resources using semantic web technologies.
• Employed backward chaining inference to identify and explain DDI mechanisms.
• Evaluated the framework with a case study involving irinotecan and levofloxacin.

Main Results:

• The framework successfully predicted and explained DDI mechanisms for the drug pair.
• Demonstrated the ability to interrogate and describe interaction mechanisms from integrated data.
• Case study validated the framework's effectiveness in exploring potential DDIs.

Conclusions:

• The developed framework effectively identifies and explains DDI mechanisms.
• This approach can enhance drug discovery, design, and cost-effective DDI identification.
• Facilitates clinical decision-making by providing insights into DDI mechanisms and relevance.