Drug-Drug Interactions Prediction Based on Similarity Calculation and Pharmacokinetics Mechanism

Abstract

Drug-drug interactions (DDIs) are one of the major causes of adverse drug events (ADEs), therefore, the prediction of DDIs for avoiding the ADEs is an important issue, which can help medical researchers economize research related resources in clinical trials. This study aims to predict DDIs based on drug similarity and ontology reasoning, and accordingly gives some possible explanations to why these drugs have DDIs. we develop a DDIs ontology integrated with similar drugs and pharmacokinetics(PK) mechanism, and formulate rules for inferring DDIs. Our method extends the existing research ideas, not only adds extrapolation of unknown data, but also reduces reliance on known data, and innovatively combines similar drugs with PK mechanism, which proved to be useful for inferring DDIs and can give some possible explanations for these DDIs. Besides our study is less demanding for data type, and the rules are more concise.