Top-down modeling methods are based on the idea of collecting all known information on a system in a list of constraints. Rather than in one particular model this generally results in a set of models that cannot be further distinguished using the available information.

Properties shared by all models in the set can be viewed as strongly supported by the invested information. Determining distinguishing characteristics for model sets can help to identify system traits that need to be clarified by further experiments. Within this project, we employ this approach in the context of logical modeling methods in systems biology.

This abstract modeling framework has proved its use in application and also allows exploitation of formal verifictaion techniques such as model checking methods. The focus is on developing approaches for extracting meaningful characteristics of model sets and dependencies between them exploitable for experimental design as well as on providing efficient implementations. Close cooperation with partners from the medical sciences working on signaling networks motivates and guides the theoretical work.