Dynamic network topology changes as result of cellular stress

The topology of metabolic or protein-protein interaction networks has been an extensively studied subject. Our primary interest on the topic is how the context, such as changing physiological state of the system, affects the network topology and connectivity within- and between the cellular functional modules and through this obtain better understating about the control mechanisms of biological systems. In this study we investigated the changes in network structure as results of oxidative stress. We collected S. cerevisiae data on protein-protein interactions (DIP), metabolic pathways (KEGG), gene regulatory relationships (TRANSFAC) into our bioinformatics system [1]. We collected available experimental gene expression data from S. cerevisiae during oxidative stress response at different time points [2]. The data was integrated into the network context by defining criteria for evaluating presence or absence of proteins in the integrated network. Swissprot index of S. cerevisiae proteins [3] was used to translate ORF identifiers into the expression dataset to Swissprot protein accession numbers. We thus reduced the networks and reconstructed condition specific networks corresponding to each expression data by removing all the proteins that are absent and their incident links. Structural organization of these networks was compared by studying topological characteristics. We found that the degree distribution of most of the networks obtained was different from the power law. Additionally, we found changes in clustering coefficient, i.e. local connectivity properties, at two specific time points during the oxidative stress response. Our results suggest the connectivity of the system is being modulated as a response to stress or other external stimuli.