Modeling Polypharmacy Side Effects with Graph Convolutional Networks

Abstract

Motivation: The use of multiple drugs, termed polypharmacy, is common to treat patients with complex diseases or co-existing medical conditions. However, a major consequence of polypharmacy is a much higher risk of adverse side effects for the patient. Polypharmacy side effects emerge because of drug-drug interactions, in which activity of one drug may change, favorably or unfavorably, if taken with another drug. The knowledge of drug interactions is often limited because these complex relationships are rare, and are usually not observed in relatively small clinical testing. Discovering polypharmacy side effects thus remains an important challenge with significant implications for mortality and morbidity of patients. Results:Here we take advantage of the increasing understanding of network-based drug activity to introduce Decagon, an approach for modeling polypharmacy side effects. The approach constructs a multimodal graph of protein-protein interactions, drug-protein target interactions, and the polypharmacy side effects, which are represented as drug-drug interactions, where each side effect is an edge of a different type. Decagon is developed specifically to handle such multimodal graphs with a large number of edge types. Our approach develops a new graph convolutional neural network for multirelational link prediction in multimodal networks. Unlike approaches limited to predicting simple drug-drug interaction scalar values, Decagon can predict the exact side effect, if any, through which a given drug combination manifests clinically. Decagon accurately predicts polypharmacy side effects, outperforming baselines by up to 69%. We find that it automatically learns representations of side effects indicative of co-occurrence of polypharmacy in patients. Furthermore, Decagon models particularly well polypharmacy side effects that have a strong molecular basis, while on predominantly non-molecular side effects, it achieves good performance because of effective sharing of model parameters across edge types. Decagon creates an opportunity to use large molecular, pharmacological, and patient population data to flag and prioritize polypharmacy side effects for follow-up analysis via formal pharmacoepidemiological studies.Availability: Source code and preprocessed datasets are at http://snap.stanford.edu/decagon.

Graph convolutional networks: a comprehensive review

“…In drug discovery, DeepChemStable [113], an attention-based graph convolutional network mode, is used for chemical stability prediction of a compound. Besides, by modeling the protein-protein interactions, drugprotein target interactions into a multimodal graph, graph convolutions can be applied to predict polypharmacy side effects [114]. Another important application in chemistry is the molecular property prediction.…”

Discovery of inflammatory bowel disease-associated miRNAs using a novel bipartite clustering approach

“…Biological network analysis of the pairwise combinations of protein, miRNA, metabolite, conserved functional subsequences, and factor binding sites can predict or understand different cellular mechanisms. Graph convolutional and deep learning methods are also popular technique on prioritizing or predicting the outcome of a gene or disease from such network [9][10][11]. In the current work, we mainly focused on MRM detection from MTIs by a new biclustering approach we recently developed [12,13].…”

Ontologies, Knowledge Representation, and Machine Learning for Translational Research: Recent Contributions

“…DeepWalk and extensions of DeepWalk such as node-2vec provide class or multiclass predictions for nodes, in which every node in the graph is assigned one or more labels representing a finite set of categories [28,29]. Graph embedding techniques have been used for a number of prediction tasks in the biomedical domain including the prediction of polypharmacy side effects [30].…”