Highlight talk – Theme: Systems
Abstract
Characterizing drug efficacy and safety has been an attractive but challenging task for the computational biology community. Many methods have attempted to identify novel drug uses and drug adverse effects, yet they rely on the similarity between drugs and diseases to make predictions, hindering both their interpretability and ability to discover novel associations. We proposed a network-based “proximity” measure to quantify the closeness between drugs, diseases and side effects and showed that proximity could explain the drug’s effect on a given phenotype. Our analysis highlighted efficacy issues for drugs used in parkinsonian and several inflammatory disorders and provided insights for several recent clinical failures. We also demonstrated that proximity could accurately predict drug side effects. The key advantage of proximity over existing similarity based approaches is that it is blind to known drug-disease or drug-side effect associations and it provides interactome-level evidence towards the drug’s therapeutic and adverse effect.
Authors
Emre Guney, Institute for Research in Biomedicine (IRB Barcelona), Spain
Jorg Menche, Research Center for Molecular Medicine of the Austrian Academy of Sciences, Austria
Marc Vida, Dana Farber Cancer Institute, United States
Albert-Laszlo Barabasi, Northeastern University, United States
Source of publication
Network-based in silico drug efficacy screening, 2016, Nature Communications, 7:10331