An algorithmic framework for predicting side-effects of drugs

One of the critical stages in drug development is the identification of potential side effects for promising drug leads. Large scale clinical experiments aimed at discovering such side effects are very costly and may miss subtle or rare side effects. To date, and to the best of our knowledge, no computational approach was suggested to systematically tackle this challenge. In this work we report on a novel approach to predict the side effects of a given drug. Starting from a query drug, a combination of canonical correlation analysis and network-based diffusion are applied to predict its side effects. We evaluate our method by measuring its performance in cross validation using a comprehensive data set of 692 drugs and their known side effects derived from package inserts. For 34% of the drugs the top scoring side effect matches a known side effect of the drug. Remarkably, even on unseen data, our method is able to infer side effects that highly match existing knowledge. Our method thus represents a promising first step toward shortcutting the process and reducing the cost of side effect elucidation.