Функционална анотација во протеински интеракциски мрежи

Abstract

Complex networks have recently become the focus of research in many fields. Their structure reveals crucial information for the nodes, how they connect and share information. In this thesis protein interaction networks are analyzed as complex networks from two aspects, how to extract knowledge about the functions of an unknown protein from the network in a direct manner and how to utilize the network’s functional modular structure to achieve the same goal. Different graph representations for the protein interaction network are proposed, each having different level of complexity and different inclusion of the annotation information within the graph and each of these is explored as to what the benefits and the drawbacks are when it is used in the functional annotation process. The first research direction is based on the hypothesis that the simultaneous activity of sometimes functionally diverse functional agents comprises higher level processes in different regions of the protein interaction network. In line with this a functional neighborhood is defined and constructed by using random walks on the protein interaction graph. Modularity is utilized via clustering in the protein interaction graph with the purpose of obtaining functionally enriched protein groups that can later be used in the functional annotation of an unknown protein. The experiments are performed using a purified and reliable Saccharomyces cerevisiae protein interaction network, which is then used to generate the different graph representations. We evaluate results in regards of biological validity and function prediction performance. Our results indicate that the new complex graph representations improve the prediction process and the proposed algorithms are better or in line with the up to date best algorithms as referenced in the literature.