Speaker: Robert Clewley
Abstract. Engineers have long investigated unknown systems by embedding them in test environments that are designed to mimic and control aspects of their original environment. This technique is increasingly applicable to scientific investigation of natural systems having modular and hierarchical structures, and to the mathematical models derived from them.
With the goal of reverse engineering models from data or forward engineering models from hypotheses, we will consider a novel use of hybrid dynamical systems integrated with model reduction methods, using new computational infrastructure. The hybrid systems permit an investigator to construct "scaffolding" around any module of interest, allowing a multi-tiered reduction of a complex model that employs different levels of abstraction. This creates new opportunities for qualitative optimization of models, and for a new view of optimally parsimonious descriptions of biological mechanisms.