System-level understanding of the relationships between drugs and targets is very important
for enhancing drug research, especially for drug function repositioning. The experimental methods
used to determine drug-target interactions are usually time-consuming, tedious and expensive,
and sometimes lack reproducibility. Thus, it is highly desired to develop computational methods for
efficiently and effectively analyzing and detecting new drug-target interaction pairs. With the explosive
growth of different types of omics data, such as genome, pharmacology, phenotypic, and other
kinds of molecular networks, numerous computational approaches have been developed to predict
Drug-Target Interactions (DTI). In this review, we make a survey on the recent advances in predicting
drug-target interaction with network-based models from the following aspects: i) Available public
data sources and benchmark datasets; ii) Drug/target similarity metrics; iii) Network construction; iv)
Common network algorithms; v) Performance comparison of existing network-based DTI predictors.
Keywords: Drug-target interaction prediction, Drug repositioning, Drug similarity metrics, Target similarity metrics, Network
construction, Network models.
Rights & PermissionsPrintExport