Getting poster data...
Corinna Vehlow, Daniel Weiskopf (Universität Stuttgart, Visualisierungsinstitut (VISUS), Allmandring 19, 70569 Stuttgart)Many biological problems are modeled as graphs and analyzed using biological networks techniques. Similar to other application domains of graphs, networks are not regular but consist of communities, i.e., highly interconnected sets of vertices. In protein-protein-interaction networks, e.g., such structural subunits present protein complexes. However, communities in PPI networks do overlap, since many proteins can play several distinct roles in different contexts. We have developed a network visualization techniques for fuzzy overlapping communities that allows users to analyze communities of proteins and their overlapping nature at different levels of detail and to identify to what extent shared (fuzzy) proteins are involved in particular protein complexes. Besides the network topology itself, it is often of interest to analyze the multivariate attributes of relations. Fluxes or parameter in metabolic networks derived using parameter sampling, e.g., are often correlated among each other. To analyze such correlations between edge attributes, we have developed a visualization for edge-edge-relations.