The Next Generation of Machine Learning in DDIs Prediction

Area of Science:

• Pharmacology
• Computational Biology
• Bioinformatics

Background:

• Drug-drug interactions (DDIs) can lead to severe adverse events like cardiotoxicity and hepatotoxicity.
• Traditional drug assays are time-consuming, expensive, and insufficient for the rapidly growing number of new drugs.
• Computational methods offer a promising alternative for predicting DDIs.

Purpose of the Study:

• To review the literature on computational methods for predicting drug-drug interactions.
• To highlight the application of deep learning models in DDI prediction.
• To discuss current challenges and future opportunities in computational DDI prediction.

Main Methods:

• Literature review of computational techniques for DDI prediction.
• Focus on deep learning models applied to DDI prediction.
• Analysis of widely used datasets in DDI research.

Main Results:

• Deep learning models show significant potential for predicting drug-drug interactions.
• The review covers various datasets and state-of-the-art deep learning architectures.
• Identified challenges and opportunities for computational DDI prediction.

Conclusions:

• Computational approaches, particularly deep learning, are essential for efficient and accurate DDI prediction.
• Further research is needed to address challenges and leverage opportunities in this field.
• Advancing computational DDI prediction can improve drug safety and development.