External optimal control of self-organisation dynamics in a chemotaxis reaction diffusion system.

Abstract

Detailed quantitative understanding and specific external control of cellular behaviour are general long-term goals of modem bioscience research activities in systems biology. Pattern formation and self-organisation processes both in single cells and in distributed cell populations are phenomena which are highly significant for the functionality of life, because life requires to maintain a highly organised spatiotemporal system structure. In particular chemotaxis is crucial for various biological aspects of intercellular signalling and cell aggregation. As an example for model based control of self-organising biological systems, we describe numerical optimal control of E. coli bacterial chemotaxis based on a 1-D two-component partial differential equation (PDE) model of reaction diffusion type. We present a numerical scheme to force cell aggregation patterns to particular desired results by applying a boundary influx control of chemoattractant without interfering with the system itself. Optimal controls are numerically computed by using a specially tailored interior point optimisation technique applied to a direct collocation discretisation of the control function and the PDE constraint. The objective to be minimised is the deviation of a desired cell distribution from the cell density, which results from the dynamics of the controlled system.