Complex diseases are commonly believed to be caused by the breakdown of several correlated genes rather than individual genes. The availability of genome-wide data of high-throughput experiments provides us with new opportunity to explore this hypothesis by analysing the disease-related biomolecular networks, which are expected to bridge genotypes and disease phenotypes and further reveal the biological mechanisms of complex diseases. In this study, the authors review the existing network biology efforts to study complex diseases, such as breast cancer, diabetes and Alzheimer's disease, using high-throughput data and computational tools. Specifically, the authors categorise these existing methods into several classes based on the research topics, that is, disease genes, dysfunctional pathways, network signatures and drug-target networks. The authors also summarise the pros and cons of those methods from both computation and application perspectives, and further discuss research trends and future topics of this promising field.