ABSTRACT
The cell brings various proteins in specific contexts with respect to internal and external perturbations to invoke appropriate responses. The context of a protein could be immune response against invading pathogens or a metabolic pathway in which a set of proteins breaks down glucose molecules to provide energy. Therefore, the contextual information of a protein helps us in understanding its function at local and global level. The last two decades have witnessed a significant progress in identifying protein context and understanding protein organization at systems level. The context of proteins is conceptualized in the form of a network or a graph where proteins that participate in related functions are connected by edges. The global network can be derived by assembling contextual information of all proteins encoded by a cell and provides a perspective on the functioning of proteins in context of others. Several methods have been proposed to infer contextual information of a protein. Genomic context based methods assume that the co-evolutionary signals and neighborhood of protein encoding genes within the genome sequences reflect their functional dependence. Co-expression of protein coding genes also turned out to be a hallmark of their linked roles in specific functions. These methods have been providing a wealth of information on the organization of pathways at the global level, on functional clues for uncharacterized proteins based on their connectivity with known proteins, and identification of disease related proteins and selectively targeting drugs. This chapter reviews the computational protein-protein functional linkage prediction methods developed in the post-genomic era. The chapter concludes with the proposal of a gene co-regulation based, novel method for functional linkage prediction.
