Insights into drug adverse reactions prediction through Mendelian randomization: a review

Abstract Adverse drug reactions pose a significant threat to patient safety and public health and often become apparent only after widespread clinical use. Mendelian randomization (MR) analysis is a valuable tool that can be used to infer causality by using genetic variants as instrumental variables, which can predict the occurrence of adverse drug reactions before they occur. Compared with traditional observational studies, MR Analysis can reduce the potential bias of confounding factors. This article reviews the principles of MR Analysis and its application in the prediction of adverse drug reactions, the challenges and future directions, and summarizes how to harness the power of this innovative epidemiological method to put us at the forefront of improving drug safety assessment and personalized medicine.