A survey of current integrative network algorithms for systems biology

Abstract

The goal of systems biology is to gain a more complete understanding of biological systems by viewing all of their components and the interactions between them simultaneously. Until recently, the most complete global view of a biological system was through the use of gene expression or protein-protein interaction data. With the increasing number of high-throughput technologies for measuring genomic, proteomic, and metabolomic data, scientists now have the opportunity to create complex network-based models for drug discovery, protein function annotation, and many other problems. Each technology used to measure a biological system inherently presents a limited view of the system. However, the combination of multiple technologies can provide a more complete picture. Much recent work has studied integrating these heterogeneous data types into single networks. Here we provide a survey of integrative network-based approaches to problems in systems biology. We focus on describing the variety of algorithms used in integrative network inference. Ultimately, the survey of current approaches leads us to the conclusion that there is an urgent need for a standard set of evaluation metrics and data sets in this field.