Hierarchical Modeling for Computational Biology

Diverse hierarchies play a role in modeling and simulation for computational biology, e.g. categories, abstraction hierarchies, and composition hierarchies. Composition hierarchies seem a natural and straightforward focus for our exploration. What are model components and the requirements for a composite approach? How far do they support the quest for building blocks in computational biology? Modeling formalisms provide different means for composing a model. We will illuminate this with DEVS (Discrete event systems specification) and the pi calculus. Whereas in DEVS distinctions are emphasized, e.g. between a system and its environment, between properties attributed to a system and the system itself, these distinctions become fluent in the compact description of the pi calculus. However, both share the problem that in order to support a comfortable modeling, a series of extensions have been developed which also influence their possibility to support a hierarchical modeling. Thus, not individual formalisms but two families of formalisms and how they support a composite modeling will be presented. In computational biology one type of composite model deserves a closer inspection, as it brings together the wish to compose models and the need to describe a system at different levels in a unique manner, i.e. multi-level models.