Team:ETHZ Basel/Achievements/Systems Design

From 2010.igem.org

Revision as of 12:47, 27 October 2010 by Georgerosenberger (Talk | contribs)

Systems Design

Schematical overview of the connection between biology and engineering. Wet laboratory provided mathematical modeling and information processing with experimental data and knowledge, while mathematical modeling supported the other two with parameter evaluation. Information processing used the combined mathematical model as test bench.

E. lemming is a special project, because the final product has biological and engineering aspects in equal parts. For this reason, it was decided to connect the wet laboratory and the modeling subteam closely from the very beginning to make the best decisions for the whole project.

Wet Laboratory: Experimental data

Bacterial movement

Besides its theoretical importance of closing the modeling loop of our system, the bacterial movement model also brings together biological and theoretical results, by the in silico reproduction of the in vivo observed data.

All parameters of this model are based on statistical estimates of the biologically observed chemotaxis behavior. Furthermore, by fixing the parameters for which the biological evidence supports independence, the model is accurately estimating the remaining set of parameters, providing us with reliable biological feedback & novel insights.

In this way, the wet laboratory team received the modeler's input on the predicted final behavior of E. lemming, which served as a further input in tuning the design of the wet laboratory experiments.

Cell detection

The starting point of the cell detection algorithm were the microscope images of wild type E. coli, which are targeted for detection. The parameters of the algorithm (eg. threshold for Sobel filter, filter parameters to distinguish real cells from other objects & spots) were tuned using the properties of real cells.

Mathematical Modeling: Experimental design

Objective function for parameter evaluation. Evaluation to achieve modeling insights relies on optimization of the relative amplitude of the regulating protein for the tumbling / directed movement ratio.

Modeling insights for wet laboratory

Mathematical modeling of biological pathways should never be an end in itself. Instead, mathematical models should be build to give insights into the pathway of interest, which are otherwise hard to obtain, or to speed up experimental work. In this project, a molecular setup for implementation by the wet laboratory was evaluated and resulted in an effort alleviating priority list of BioBrick candidates. Experimental Design: Insights for wet laboratory provide more information about this topic.

Table 1: Evaluation results for wet laboratory
Che LSP1 LSP2 [Asp] [AP] [anchor]
CheY PIF3 PhyB 10^-6 μM 40 μM 50 μM

Modeling insights for information processing

Furthermore, the combined model was used to estimate the reaction time of the network on the red and far-red light pulses and the optimal configuration of the light pulse setup, crucial for the controller design. See Experimental Design: Insights for information processing for more details.

Table 2: Evaluation results for information processing
model tc RL tc FRL
Spiro et al. 0.1524s 0.4092s
Mello & Tu 0.1305s 0.2985s

Download

The evaluation script for experimental design can be downloaded from SourceForge.net. It requires the E. lemming Matlab Toolbox to be installed. The script is also included in the complete package.

  • [http://sourceforge.net/projects/ethzigem10/files/experimentalDesign.zip/download Download experimentalDesign from SourceForge.net]

Information Processing: Experimental validation

Information Processing is used to validate the experimental design and implementation. Using the in silico testbench from mathematical modeling, information processing was able to validate the general principle of the selected experimental design, given validity of used assumptions.