Direkt zum InhaltDirekt zur SucheDirekt zur Navigation
▼ Zielgruppen ▼

Humboldt-Universität zu Berlin - IRI Life Sciences

Humboldt-Universität zu Berlin | IRI Life Sciences | Events | Tutorials | Past events | Tutorial in Life Sciences - Jens Timmer & Clemens Kreutz: "From Data to Models"

Tutorial in Life Sciences - Jens Timmer & Clemens Kreutz: "From Data to Models"

Lecture, Workshop and Tutorial by Prof. Jens Timmer & Dr. Clemens Kreutz (Universität Erfurt)
What
  • Past
  • Tutorial
When Nov 25, 2016 from 09:30 AM to 04:00 PM (Europe/Vienna / UTC100) iCal

Lecture „Uncertainty Analysis in Systems Biology“ by Jens Timmer 

 
In biology, typically, cellular processes as signal transduction, gene regulation and metabolism are presented by graphic „cartoons“ which are static, qualitative, and descriptive. One goal of Systems Biology is to transform this presentations into dynamic, quantitative, and predictive mathematical models, typically ordinary differential equations. To this aim, parameters of the differential equations have to be estimated from time-resolved experimental data by minimizing some objective function. This comes with at least four types of uncertainties: (i) uncertainty about finding the global minimum, (ii) uncertainty of the estimated parameters due to uncertainty in the data, (iii) uncertainty of model predictions due to uncertainty of the estimated parameters, (iv) uncertainty about the model structure. We discuss reasons for and a simple procedure to deal with uncertainty (i), show how the profile likelihood can deal with uncertainties (ii) and (iii), and will briefly touch uncertainty (iv). 
 
11:00 a.m. 
Workshop „Problems in parameter estimation“
Workshop with the presentation and discussion of selected problems from participants. Please submit your abstract until Nov 15 to stefanie.scharf@iri-lifesciences.de.
 
1:00 p.m. 
Tutorial „Data2Dynamics“ by Clemens Kreutz 
In this tutorial the Data2Dynamics modelling framework is introduced. Parameter estimation, identifiability analyses and calculation of prediction uncertainties are illustrated. The software requires Matlab and is available at www.data2dynamics. org.