Dr. Sorin Mitran
October 17, 2016
Colloquium by Dr. Sorin Mitran, University North Carolina
Title: "Deterministic and stochastic reduced order modeling of microscopic organism motility mechanisms"
Microscopic organisms exhibit various modes of propulsion: ciliary or flagellar beating, lamellipodium protrusion followed by attachment/detachment to a substrate, taking over control of actin production in a host cell. High-throughput computational simulation can provide detailed description of specific motility aspects, but their cost and complexity is an impediment to furthering biological understanding of organism propulsion. This talk presents work on the transformation of detailed computational simulation into tractable reduced-order models. Two specific cases are considered: (1) ciliary propulsion to illustrate model reduction of a deterministic system, and (2) propulsion of Listeria monocytogenes to illustrate aspects of stochastic model reduction. In ciliary propulsion, molecular dynamics level computation is used to furnish a detailed description of mechanical behavior of the microtubule constituents of a cilium. The data is used to construct a finite element model that is markedly different from the Euler-Bernoulli beam models typically used in cilia studies. L. monocytogenes moves by taking over the production of actin within a host cell. Stochastic modeling of the growth of the host cytoskeleton catalyzed by L. monocytogenes is used to construct a statistical model of the flight/forage behavior that can be used to infer infection virulence. Model reduction in this case involves consideration of the differential geometry of probability distributions, a field of study known as information geometry. The model reduction procedures are presented at a conceptual level, avoiding technical details, concentrating on the goal of arriving at correct models of direct utility to biology.