Sensitivity analysis of a whole cell model
|has title::Sensitivity analysis of a whole cell model|
|Student name:||student name::Rui Peng Wang|
|Second reader:||has second reader::Sanne Abeln|
|Company:||has company::Systems Bioinformatics IBIVU|
Background Optimization processes in microorganisms are of interest to scientists and industry. To improve biomass yield efficiency it is necessary to optimize microorganisms. In our project, we would like to understand how microorganisms optimize themselves, and whether we can explain it by tuning of protein levels. For that purpose, we have built a whole-cell model, in which important metabolic pathways are represented by modules . In particular, we would like to comprehend the interplay of different modules in the cell to achieve optimal states.
Approach The model parameters are taken form enzyme databases, which are based on data from experiments conducted in various conditions, which may be far away from in vivo conditions. Therefore, it is of importance to see how sensitive our model is to the parameter values used. Furthermore, to have an insight for the interplay between metabolic actors, it is of great value to see how robust is the model. Therefore we will be conducting sensitivity analysis on this whole-cell model. The sensitivity analysis is based on Monte Carlo simulation, which allows analysis of multiple parameters simultaneously. Sensitivity analysis results will show how the solution space looks like, and which parameters matter the most. This can make us see how sensitive actual cells are to these parameters, and will push us to think about what that could mean in the course of evolution.
The procedure will be automatised so that the optimization platform can perform iterations. The Monte Carlo procedure will iterate through a list of sampled kinetic parameters. This procedure will be implemented in MATLAB.
Timeline The internship project will take place in the Systems Bioinformatics group of the Institute for Integrative Bioinformatics VU (IBIVU) in Amsterdam. The internship duration is ~4 or 5 months starting in April 2011.
April: Preparing, reading, building model, analysing
May: implementing monte carlo method, analysing
June: Analysing, start writing report
July: Analysing, writing report
August: writing report
 Shifts in metabolic efficiency. D. Molenaar et al. 2009 EMBO and Macmillan Publishers Limited.
 Dynamic modelling and analysis of biochemical networks: mechanism-based models and model-based experiments. N. A.W. van Riel. 2006. Oxford University Press. Briefings in Bioinformatics Vol 7. NO 4. 364-374.