Proteins are the main catalysts, structural elements, signalling messengers and molecular machines of biological tissues [1
]. Protein–protein interactions (PPIs) are extremely important in orchestrating the events in a cell. They form the basis for several signal transduction pathways in a cell, as well as various transcriptional regulatory networks. The availability of complete and annotated genome sequences of several organisms has led to a paradigm shift from the study of individual proteins in an organism to large-scale proteome-wide studies of proteins, which interact in a beautifully concerted network of metabolic, signalling and regulatory pathways in a cell. In general, the behaviour of a system is quite different from merely the sum of the interactions of its various parts. As Anderson put it as early as 1972, in his classic paper by the same title, "More is different
] — it is not possible to reliably predict the behaviour of a complex system, despite a good knowledge of the fundamental laws governing the individual components. Comparative genomics at a primary sequence level has also indicated that species differences are due more to the difference in the interactions between the component proteins, rather than the individual genes themselves [3
]. Consequently, several efforts have been made to identify these interactions, in an attempt to understand biological systems better [4
]. The need to understand protein structure and function has been a critical driving force for biological research in the recent decades. With the advent of high-throughput experiments to identify PPIs, more knowledge on protein function has been obtained, together with the development of several methods to predict and study the interactions between proteins.
A wide variety of methods have been used to identify protein–protein associations; these associations may range from direct physical interactions inferred from experimental methods to functional linkages predicted on the basis of computational analyses. In the past, experimental methods based on microarrays and yeast two-hybrid, as well as computational methods based on protein sequences and structures have been developed and widely used. Given the difficulties in experimentally identifying PPIs, a wide range of computational methods have been used to identify protein–protein functional linkages and interactions. These methods range from identifying a single pair of interacting proteins at one end, to the identification and analysis of a large network of thousands of proteins, the latter as large as that of an entire proteome of a given cell.