In biological systems, most processes are carried out through orchestration of multiple interacting molecules. These interactions are often abstracted using network models. A key feature of cellular networks is their modularity, which contributes significantly to the robustness, as well as adaptability of biological systems. Therefore, modularization of cellular networks is likely to be useful in obtaining insights into the working principles of cellular systems, as well as building tractable models of cellular organization and dynamics. A common, high-throughput source of data on molecular interactions is in the form of physical interactions between proteins, which are organized into protein-protein interaction (PPI) networks. This chapter provides an overview on identification and analysis of functional modules in PPI networks, which has been an active area of research in the last decade.
Proteins that make up a functional module tend to interact with each other and form a densely connected subgraph in a PPI network.Motivated by this observation, module identification is often formulated as a problem of partitioning a PPI network into dense subgraphs, which is also known as graph clustering. This chapter begins with a brief introduction to the module identification problem in PPI networks. Then, graph theoretical measures of modularity such as density, clustering coefficient and edge connectivity are introduced. Algorithmic approaches for identifying modules are then presented in a systematic manner. These clustering approaches are broadly categorized as (i) Bottom-up (ii) Top-down (iii) Iterative Improvement and (iv) Flow Based methods. Subsequently, a sample application of modularization, namely, predicting the function of uncharacterized proteins, is briefly discussed. More advanced methods to identify functional modules often integrate other data sources such as gene expression data with PPI data or use multiple networks to find conserved regions in the networks. After an overview on these advanced methods, some exercises are presented to the reader.