Reasoning over networks with BioRevise and PowerGraphs

The analysis of protein interactions is a fundamental application of bioinformatics. Previous deliverables discussed the BioCham pathway analysis engine developed in the Paris group, this deliverable introduces the BioRevise systems, which integrates a non-monotonic reasoning engine orginially developed in Lisbon with the application of reasoning over metabolic pathways. We discuss examples, modelling of the networks, reasoning, and the demonstrator. Complex networks such as the metabolic networks in BioRevise require sophisticated visualisation. We develop power graph analysis, which identifies modules in networks and visualises them in a compact fashion. The theoretical basis for power graphs are the identification of bicliques and cliques in graphs. Power graphs can be applied to any network including class dependencies and UML diagrams as discussed in groups I1 and I3. The deliverable introduces an algorithm and discusses a number of bioinformatics applications for power graphs. Keyword List Reasoning, revision, conflic resolution, metabolic network, protein interaction, visualisation Project co-funded by the European Commission and the Swiss Federal Office for Education and Science within the Sixth Framework Programme.