Accessible Protein Interaction Data for Network Modeling. Structure of the Information and Available Repositories

In recent years there has been an incredible explosion of computational studies of molecular biology systems, particularly those related to the analysis of the structure and organization of molecular networks, as the initial steps toward the possible simulation of the behavior of simple cellular systems. Needless to say, this task will not be possible without the availability of a new class of data derived from experimental proteomics. Large-scale application of the yeast two-hybrid system, affinity purification (TAPs-MS), and other methodologies are for the first time providing overviews of complete protein interaction networks. Interestingly a number of computational methods are also contributing substantially to the identification of protein interactions, by comparing genome organization and evolution. Other disciplines, such as structural biology and computational structural biology, are complementing the information on interaction networks by providing detailed molecular descriptions of the corresponding complexes, which will become essential for the direct manipulation of the networks using theoretical or experimental methods. The storage, manipulation and visualization of the huge volumes of information about protein interactions and networks pose similar problems, irrespective of the source of the information: experimental or computational. In this sense, a number of competing systems and emerging standards have appeared in parallel with the publication of the data. In this review, we will provide an overview of the main experimental, high-throughput methods for the study of protein interactions, the parallel developments of computational methods for the prediction of protein interactions based on genome and sequence information, and the development of databases and standards that facilitate the analysis of all this information.