Biological systems are inherently modular but the modules are highly connected, with multiple feedbacks. Automated reconstruction from archived modules requires adherence to standards, the use of ontologies, and computer-recognition of relationships between processes. Even more challenging is the automation of the processes of reducing model complexity for using modules in higher level, more speedily computable versions. Success in simplification means a reduction in robustness, defining robustness as the generic ability to respond appropriately to changes, thus recognizing that only certain types of responses can be preserved while reducing model complexity. The residual fundamental problem then is, when inadequacy is identified, how to unreduce the model efficiently to reincorporate only the critical missing characteristics.

Back to Minisymposium_planning.

ABSTRACTS

• Beard:Merging Cellular Biochemical Models Using Physicochemically Rigorous Rules

Daniel A. Beard, Ranjan K. Dash, Feng Qi, Kalyan C. Vinnakota, and Fan Wu

Biotechnology and Bioengineering Center and Department of Physiology, Medical College of Wisconsin, Milwaukee, WI 53226

Simulations of cellular systems are optimally realistic and meaningful only when appropriate physical chemical rules are adopted. Since a great deal of information is available regarding the thermodynamic and ion-binding properties of biochemical reactants, it is possible to construct simulations of biochemical systems that properly incorporate these data. Specifically, realistic simulations of biochemical systems require accounting for: (i.) the complex multiple equilibria of biochemical species and dynamic buffering of ions; and (ii.) the pH and ionic dependence of enzyme kinetics and apparent equilibria and thermodynamic driving forces for biochemical reactions. Using a formal method that treats these phenomena, we can develop and validate computational models of systems of unprecedented complexity. In addition, rigorous physical chemical rules facilitate non-ambiguous model integration while reducing uncertainty in parameter estimates and improving the reliability of model predictions. As an example we will show how a model of cardiac energy metabolism that tracks more than 100 species is developed, parameterized, validated, and used to generate hypotheses and understand emergent phenomena.

• V.A. Shiva Ayyadurai, C. Forbes Dewey, Jr.:Dynamic Integration of Distributed Biological Pathway Models Using Cytosolve

Dept of Biological Engineering, Massachusetts Institute of Technology, Cambridge MA 02138

One prevailing strategy to model the whole cell is a bottom’s up one: first integrate smaller biological pathway models to build larger models of cellular function and then link such larger models to produce a computational model of the whole cell. The current approach to realize this strategy is a monolithic one, where the model integrator manually merges the computer source codes of smaller biological pathway models to create one large monolithic computer program containing the merged source codes. The monolithic approach is not scalable for integrating the thousands of biological pathway models necessary to model the whole cell. We present Cytosolve, a new method that dynamically integrates a distributed ensemble of biological pathway models without the need to merge the source codes of the individual biological pathway models. Cytosolve allows each biological pathway model to reside at its own location on any computer worldwide, where the authors of each model can independently maintain and update the model’s source code. As more biological pathway models develop in a disparate and decentralized manner, Cytosolve provides a scalable method to integrate complex models of cellular function, and eventually to model the whole cell. In this talk, we will present the Cytosolve architecture and share two working examples: (1) integrating the EGFR pathway of Kholodenko, and (2) an integrative model of the IFN response to viral infection.

• Rod Smallwood:Robust large-scale individual-based modelling in biology

University of Sheffield - agent-based cell models (r.smallwood@sheffield.ac.uk)

We are interested in the complex (emergent) behaviour of large numbers (millions) of interacting entities. The interactions might take place on the same length and time scale or on a hierarchy of length and time scales. The entities may be physical objects which are constrained to obey the laws of physics, and may have complex networks embedded within them. Examples include cell signalling networks (molecule level), tissue development, wound healing, and the development of malignancy (cell level), and the behaviour of ant colonies and macro-economic behaviour in human societies (society level). Our aim is to understand behaviour and predict the effect of intervention, in order to design therapeutic interventions or influence social policy. Properly engineered and robust software tools are essential if the goal is to use the software to inform clinical and policy decisions. We are developing FLAME (Flexible Large-scale Agent Modelling Environment, http://flame.ac.uk) to provide these robust software tools. The presentation will provide an overview of our approach to modeling complex systems and the aims of the software development team, and examples of model building using the software, including links to physics models and other modeling paradigms (e.g. COPASI).

Media:SIAM_2008_Smallwood.ppt