Skip to content. | Skip to navigation

Personal tools
Sections

Carsten Maus, Mathias John, Stefan Rybacki, and Adelinde M Uhrmacher (2010)

Towards Rule-Based Multi-Level Modeling

Poster, Edinburgh, UK, 11th International Conference on Systems Biology (ICSB).

Rule-based approaches help to reduce model complexity for modeling dynamic systems by defining structured objects and rules that control alteration of and interaction between these objects.  Only the basic model entities and relevant reactions have to be described.  Thereby, rule-based modeling helps to effectively handle the problem of combinatorial explosion which might easily occur in todays systems biology projects.  Furthermore, its intuitive modeling metaphor that is along the lines of well-known notations of (bio-)chemical reactions might be an additional reason why rule-based modeling has become subject of increasing attention during the past years. Existing popular rule-based approaches like BioNetGen and Kappa do not provide explicit support for describing systems at multiple levels of abstraction.  However, although this has proofed many times to be sufficient, in order to understand certain biological systems additional means for structuring the model might be desirable.  For example, besides molecules in a well-mixed solution, one might wish to model certain molecules and reactions spatially constrained and thus at a higher level of detail.  Another example and maybe the most obvious motivation for multi-level modeling comes naturally with the hierarchical organization of life.  Biological systems are composed of objects at different organizational levels, e.g. proteins form cells which are further grouped to tissues and so on.  How to combine such different levels within the same model and how they can interact with each other are crucial questions for any multi-level approach. We developed a formalism that explicitly supports multi-level modeling for biological systems.  Nested entities with arbitrary attributes can be used to mimic organizational abstractions from tissue to cells to molecules. Due rule schemas and expressions that work on attributes and constrain reaction rates, models are kept small and relating different abstraction or composition levels is facilitated.  As shown, even spatial dynamics can easily be realized.

Document Actions