Some studies on protein function prediction from protein interaction network

Abstract

Proteins play a crucial role in all biological activities in living organisms. To know about the cell functions very closely, it is necessary to map and sequence the genomes of different organisms. Success of Human Genome Project aided by rapid DNA sequencing techniques, has become a landmark event in this regard. Under this project it has been possible to prepare the complete sequence of the human genome, estimated to contain about 3 billion base pairs of nucleotides in double helices of DNA molecular structures. As a result, a deluge of biological sequence data of DNA and protein has been created. Techniques from Informatics and Computer Science are applied on these high throughput sequence data to understand and organize the information about the associated molecules, viz., proteins and DNA, under the discipline of Bioinformatics. These are predictive techniques with limited accuracies but are much faster compared to existing laboratory methods. As a result, protein function determination has become an extensive research as it contributes valuable information for designing drugs of different diseases. These computational based methods are based on aspects of molecular biology such as the gene and protein sequence and structure, the gene neighborhood, gene fusions, cellular localizations and protein-protein interactions. The prediction of protein functions based on protein interaction information is an emerging area of research. Since a protein generally interacts with more than one other proteins, these interactions can be structured to form a network and hence the name Protein-Protein Interaction Network (PPIN). The fact that neighbors of a given protein are of similar function motivates us to predict the functions of non-annotated proteins with the neighborhood properties in the protein interaction network.