Prediction of Adverse Drug Reactions with the help of Graph Theoretical Modeling of Drug-Target-Side Effects Models

Abstract

Despite technological progresses and improved understanding of biological systems, discovery of novel drugs is an inefficient, arduous and expensive process. Research and development cost of drugs is unreasonably high, largely attributed to high attrition rate of candidate drugs due to adverse drug reactions. Computational methods for accurate prediction of drug side effects, rooted in empirical data of drugs, have the potential to enhance the efficacy of the drug discovery process. Identification of features critical for specifying side effects would facilitate efficient computational procedures for their prediction. We created networks of drugs, side-effects and targets from the empirical data to view the structures of them. Also identified their network properties and found out that in degrees of side-effect in drug side effect network follows a scale free kind of behavior. After that the drug side effect and drug target networks were searched for the correlation between them which can facilitate the side effect prediction for novel drugs. For this degree comparison, degree distribution comparison, cosine similarity and canonical correlation methods were used. Some proved to be better than others. Drug chemical properties were also taken into account for the purpose of finding the similarity between the networks. Also prominent chemical properties were identified which correlates best with side effects. We also performed check to identify why the prediction scores can give high results for some methods by performing one random network experiment.