Computational Drug Recommendation Approaches toward Safe Polypharmacy

Abstract

Adverse drug reactions (ADRs) induced from high-order drug-drug interactions (DDIs) due to polypharmacy - simultaneous use of multiple drugs - represent a significant public health problem. Unfortunately, computational effects to facilitate future polypharmacy, particularly to assist safe multidrug prescription, are still in their infancy. We formally formulate the to-avoid and safe drug recommendation problems when multiple drugs have been taken simultaneously. We develop a joint model with a recommendation component and an ADR label prediction component to recommend for a prescription a set of to-avoid/safe drugs that will induce/will not induce ADRs if taken together with the prescription. We also develop real drug-drug interaction datasets and corresponding evaluation protocols.