Contributed Session Fri.1.MA 376

Friday, 10:30 - 12:00 h, Room: MA 376

Cluster 12: Life sciences & healthcare [...]

Model discrimination and experimental design

 

Chair: Alexandra Herzog

 

 

Friday, 10:30 - 10:55 h, Room: MA 376, Talk 1

Max Nattermann
A quadratic approximation of confidence regions

Coauthor: Ekaterina Kostina

 

Abstract:
Dealing with the task of identifying unknown quantities from a set of erroneous data, the performance of a sensitivity analysis is inevitable. Without the determination of the statistical accuracy, we are not able to make any quality statements about the estimate. Consequently the result is almost meaningless. Commonly one applies linearization techniques to determine the statistical accuracy of the solution. But particularly in highly nonlinear cases this may cause problems and linear confidence regions may not be adequate.
In this talk, we are going to present and analyze a confidence region based on a quadratic approximation. Furthermore, we demonstrate our results using applications from biology. Furthermore, we discuss the impact of the new results to optimum experimental design.

 

 

Friday, 11:00 - 11:25 h, Room: MA 376, Talk 2

Tanja Binder
Numerical optimization methods for significance analysis of parameters and subsets of metabolic networks

Coauthor: Ekaterina Kostina

 

Abstract:
We have developed an efficient numerical method, based on sensitivity analysis for parametric optimization problems, that can be used to identify the most important signaling pathways and the key parameters and variables in a mathematical model that is given by a system of ordinary differential equations. In the context of metabolic pathways, our approach can be used to guide experimental biologists in their choice which proteins they should measure. Mathematically, the problem results in the question how much improvement in terms of the cost function can be achieved by adding additional terms to the underlying dynamical model, i.e., whether these are to be included or not. The cost function describes the quality of the model response in comparison to process observations. After a parameter estimation for the simplest model, we can decide fast whether additional terms should be included in the model without having to re-optimize the enlarged models. We show the capability, reliability, and efficiency of our approach using complex problems from systems biology.

 

 

Friday, 11:30 - 11:55 h, Room: MA 376, Talk 3

Alexandra Herzog
Discrimination of competetive model candidates for reversals in bacterium Myxococcus xanthus

Coauthors: Regina Gente, Ekaterina Kostina

 

Abstract:
Reversals in the gram-negative bacterium M. xanthus are still poorly understood. In general the spatial relocalisation of motility proteins is assumed to determine the dynamic orientation of the cell polarity axis and hence cell reversals. The difficulty is that experimental data from fluorescence microscopy on the simultaneous localisation of the involved proteins is both rare and of a more qualitative nature. Protein dynamics are recorded individually. Correlated data is available as qualitative observations only. Simulations of the dynamics of all involved proteins are typically the only means to study the processes under investigation.
In this talk we discuss numerical optimization methods for discrimination between available deterministic and semi-stochastic models for protein localisation of the supposed predominant proteins MglA and MglB. Qualitative reconstruction of the observed characteristic dynamics and transport times for available proteins is used as discrimination criteria. The extremely sparse experimental data sets mark a special challenge of this application.
Current results and limitations of this approach are discussed.

 

  There are three major facts that should be watched out for in all payday loans in the United States. What makes Viagra Strips better than other forms of this medicine is its easiness of usage.