W1 "Conclusions despite uncertainties"Systems biology is a rapidly growing field, which has as its key component to integrate mathematical modelling as an analysis tool for biological data. Mathematical modelling has been used extensively in many other fields, but the application to systems biology raises some new challenges, which warrants the development of new and modified theoretical methods and modelling strategies. This workshop is devoted to one such challange: how to draw conclusions despite the many uncertainties.
Data uncertainty and large mechanistically based models leads to practical unidentifiability, and this leads to the need for a differentiation of those parts of a model that can be validated/trusted (core predictions), and those parts of the model that are more or less arbitrarily uncertain because they cannot be checked by current data. This differentiation can be obtained in different ways, but we will consider at least the following options:
- Model reduction combined with practical identifiability analysis
- Generally applicable methods for model rejection (minimal modelling approaches)
- Modified optimization to determine the entire space of feasible parameters
- Set-based (or interval analysis-based) approaches giving guarantees of outer prediction/parameter boundaries
- Statistical and Bayesian approaches and their link to ordinary differential equations
With the hope of an interesting time in Linköping in June!
Email with practical info and papers to read as background (link)
Preliminary ScheduleSaturday June 19
Session on Optimization-based approaches and direct searches among the ODE-parameters
9.15-9.30 Welcome, introduction to the workshop
9.30-10.00 Gunnar Cedersund*, Linköping, Sweden, "Core-predictions - i.e. shared properties among all acceptable parameters - extends conclusive modelling beyond hypothesis testing"(background papers, 1 and 2)
10.00-10.30 Clemens Kreutz* (Andreas Raue, Jens Timmer), Freiburg, Germany, "Likelihood-based observability analysis and confidence intervals for predictions of dynamic models"
10.30-11.00 Coffee break
11.00-11.30 David Gomez-Cabrero*, Stockholm, Sweden, "Workflow for generating competing hypothesis from models with parameter uncertainty"
11.30-12.00 Dennis Rickert*, (Nicolai Fricker, Inna Lavrik, Fabian J. Theis), Munich, Germany, "Set-based reduction of a CD95 signaling model of apoptosis allows new interpretation of existing data and design of new experimental strategies" (abstract)
13.15-14.30 Group discussions
14.30-15.00 Coffee break
Session on Statistical and probabilistic approaches
15.00-15.30 Joep Vanlier*, Eindhoven, The Netherlands, "A Bayesian Approach to Predictions Despite Uncertainties"
15.30-16.00 Andrei Kramer*, Stuttgart, TBA (Bayesian approaches to parameter uncertainties)
16.00-17.00 Coffee break, final discussions on this session
19.00 Joint dinner downtown Linköping (restaurant Kniv och Gaffel)
Sunday June 19
Session on analytical, interval analysis, and set-based approaches
9.30-10.00 Jan Hasenauer*, Stuttgart, Germany, "Parameter identification, experimental design and model falsification for biological network models using semidefinite programming" (see IET Syst Biol. 2010, 4(2):119-30)
10.00-10.30 Pelle Lundberg* and Torkel Glad*, Linköping, Sweden, "Evaluation of BioSDP on realistic examples; accounting for the discretization error"
10.30-11.00 Discussion on the set-based approaches
11.00-11.30 Coffee break
11.30-12.00 Michel Kieffer*, Paris, France, "Applied Interval Analysis". Paper.
12.00-13.00 General discussion, future plans
LUNCH, END OF WORKSHOP 1
* = confirmed
- Jens Timmer, Freiburg/Linköping
- Clemens Kreutz, Freiburg
- Natal van Riel, Eindhoven
- Fianne Sips, Eindhoven/Linköping
- Joep Vanlier, Eindhoven
- Gunnar Cedersund, Linköping
- Linnea Järvstråt, Linköping
- Pelle Lundberg, Linköping
- Mikael Forsgren, Linköping
- Oscar Samuelsson, Linköping
- Rikard Johansson, Linköping
- Jacob Antonsson, Linköping
- Elin Nyman, Linköping
- David Janzén, Linköping
- Karin Lundengård, Linköping
- José Pena, Linköping
- Dennis Rickert, Munich
- Torkel Glad, Linköping
- Sina Pakazad, Linköping (probably not)
- Mohsen Alami, Linköping (probably not)
- Michael Hörnquist, Linköping
- Jan Hasenauer, Stuttgart
- Patrick Weber, Stuttgart
- Andrei Kramer, Stuttgart
- Michel Kieffer, Paris
- David Goméz-Cabrero, Stockholm
- Michel, Stockholm, (post doc, David's and Tegnér's lab)
- Ranjita, Stockholm, (Ph.D. student, David's and Tegnér's lab)
- Martin Gollvik, Linköping
Back to survey of all workshops.