Experimental implementation and validation of techniques for system identification and control of biological systemsView all posters
Telethon Institute of Genetics and Medicine (Tigem), Italy
An important branch of Control Theory, namely System Identification, aims to derive from measurement data a dynamical model of a physical system able to predict its behavior to future inputs. In biology, System Identification could be used either to design feedback control strategies to steer the biological system towards a desired goal, or even to understand the biological mechanisms underlying a biological process. However, current experimental techniques in biology allow measurements of only a few time points during a time-course, thus limiting the application of System Identification paradigms. Here, we propose an experimental approach, to make biological application of System Identification and Control Theory possible. We designed and implemented an experimental platform based on a microfluidic device, a time – lapse microscopy apparatus and, a set of automated syringes all controlled by a computer. Then we used this platform to realize in-vivo experiments on yeast cells; microfluidics allows to isolate the biological material and to precisely change cell environmental conditions (i.e. by using automated syringes to modulate the nutrients provided to yeasts). This platform allowed us to implement and compare different linear system identification methods to a simple network in S. cerevisiae, and to control gene expression within a complex network integrated in yeast cells. The ability to precisely dosage a protein in living cells could be exploited for several biological purposes. We are applying this approach to study the onset of protein aggregates involved in neurodegeneration: an engineered yeast expressing wild-type (wt) and mutated (mt) forms of the human protein alpha-synuclein under the control of the GAL1 promoter will be used to precisely control the level of wt and mt alpha-synuclein and to quantify aggregate formation propensity and protein aggregation dynamics.