Estimating chain length for time delays in dynamical systems using profile likelihood
| Autor: | Jens Timmer, Raphael Engesser, Adrian L. Hauber, J Joep Vanlier |
|---|---|
| Rok vydání: | 2019 |
| Předmět: |
Statistics and Probability
Time delays Dynamical systems theory Computer science Maximum likelihood Biochemistry Models Biological 03 medical and health sciences CLs upper limits Chain (algebraic topology) Bayesian information criterion Molecular Biology 030304 developmental biology Probability 0303 health sciences Series (mathematics) 030302 biochemistry & molecular biology Bayes Theorem Delay differential equation Computer Science Applications Computational Mathematics Computational Theory and Mathematics Research Design Identifiability Algorithm Signal Transduction |
| Zdroj: | Bioinformatics (Oxford, England). 36(6) |
| ISSN: | 1367-4811 |
| Popis: | Motivation Apparent time delays in partly observed, biochemical reaction networks can be modeled by lumping a more complex reaction into a series of linear reactions often referred to as the linear chain trick. Since most delays in biochemical reactions are no true, hard delays but a consequence of complex unobserved processes, this approach often more closely represents the true system compared to delay differential equations. In this paper, we address the question of how to select the optimal number of additional equations, i.e. the chain length. Results We derive a criterion based on parameter identifiability to infer chain lengths and compare this method to choosing the model with a chain length that leads to the best fit in a maximum likelihood sense, which corresponds to optimising the Bayesian information criterion. We evaluate performance with simulated data as well as with measured biological data for a model of JAK2/STAT5 signalling and access the influence of different model structures and data characteristics. Our analysis revealed that the proposed method features a superior performance when applied to biological models and data compared to choosing the model that maximises the likelihood. Availability Models and data used for simulations are available at https://github.com/Data2Dynamics/d2d and http://jeti.uni-freiburg.de/PNAS_Swameye_Data. Supplementary information Supplementary data are available at Bioinformatics online. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|