Abstract:
Complexes of physically interacting proteins are one of the fundamental functional units responsible for driving key biological mechanisms within the cell. Their identification is therefore necessary not only to understand complex formation but also the higher level organization and machinery of the cell. With the advent of “high-throughput” techniques in molecular biology significant amount of physical interaction data has been catalogued from organisms such as yeast, which has in turn fueled several computational techniques to systematically mine potential complexes from the network of interactions, the PPI network. In this tutorial, we will cover in detail some of the representative methods developed till date for the computational identification of protein complexes from PPI networks. We will present two insightful taxonomies of the methods that reflect how methods have evolved over the years to improve computational complex detection. We will also discuss some of the open challenges facing accurate reconstruction of complexes, the crucial ones being the presence of high proportion of errors and noise in current high-throughput datasets, and some key aspects overlooked by current complex detection methods. We hope this tutorial will provide a detailed introduction for researchers beginning to explore the areas of protein interactions and complex prediction, and also show new directions to drive further research in this area.