Application of Artificial Intelligence in Drug-Drug Interactions Prediction: A Review

Drug-drug interactions (DDI) are a critical aspectof drugresearch that can have adverse effects on patients and can lead toserious consequences. Predicting these events accurately can significantlyimprove clinicians' ability to make better decisions and establishoptimal treatment regimens. However, manually detecting these interactionsis time-consuming and labor-intensive. Utilizing the advancementsin Artificial Intelligence (AI) is essential for achieving accurateforecasts of DDIs. In this review, DDI prediction tasks are classifiedinto three types according to the type of DDI prediction: undirectedDDI prediction, DDI events prediction, and Asymmetric DDI prediction.The paper then reviews the progress of AI for each of these threeprediction tasks in DDI and provides a summary of the data sets usedas well as the representative methods used in these three predictiondirections. In this review, we aim to provide a comprehensive overviewof drug interaction prediction. The first section introduces commonlyused databases and presents an overview of current research advancementsand techniques across three domains of DDI. Additionally, we introduceclassical machine learning techniques for predicting undirected druginteractions and provide a timeline for the progression of the predicteddrug interaction events. At last, we debate the difficulties and prospectsof AI approaches at predicting DDI, emphasizing their potential forimproving clinical decision-making and patient outcomes.

Automated summary

Drug-drug interactions (DDI) are an important aspect of drug research that can have serious consequences for patients. Predicting these interactions accurately can improve clinical decision-making and treatment outcomes. Utilizing Artificial Intelligence (AI) advancements is crucial for achieving accurate forecasts of DDIs.