Computer Prediction of Adverse Drug Effects on the Cardiovascular System

Abstract

A structure–activity relationship model was constructed to predict adverse drug effects on the cardiovascular system. Models were constructed using sets of drug structures compiled by us with information about adverse effects from analyses and data integrated from various sources. The five most common cardiovascular adverse effects, i.e., myocardial infarction, ischemic stroke, cardiac failure, ventricular tachycardia, and arterial hypertension were analyzed in the work. The PASS software that has proved to be efficient in numerous studies was used to construct the appropriate models. The obtained models were accurate enough to be useful for estimating cardiovascular adverse effects of new drugs. An appropriate evaluation could be made in the very early stages of drug discovery because only the structural formula was needed for the prediction.