A hierarchical approach to model parameter optimization for developmental systems

Article (PDF Available)inBio Systems 102(2-3):157-67 · November 2010with40 Reads
DOI: 10.1016/j.biosystems.2010.09.002 · Source: PubMed
In the context of Systems Biology, computer simulations of gene regulatory networks provide a powerful tool to validate hypotheses and to explore possible system behaviors. Nevertheless, modeling a system poses some challenges of its own: especially the step of model calibration is often difficult due to insufficient data. For example when considering developmental systems, mostly qualitative data describing the developmental trajectory is available while common calibration techniques rely on high-resolution quantitative data. Focusing on the calibration of differential equation models for developmental systems, this study investigates different approaches to utilize the available data to overcome these difficulties. More specifically, the fact that developmental processes are hierarchically organized is exploited to increase convergence rates of the calibration process as well as to save computation time. Using a gene regulatory network model for stem cell homeostasis in Arabidopsis thaliana the performance of the different investigated approaches is evaluated, documenting considerable gains provided by the proposed hierarchical approach.
6 Figures
    • In order to evolve vein and leaf patterns [7, 18] , the active transport of chemicals is desirable , beyond passive diffusion. In order to improve convergence speed and escape more easily from local optima, another research topic would be the adaptation of more sophisticated algorithms, such as CMA-ES [12, 19] or the nested evolution algorithm from [14], to the evolution of reaction networks in the RDS context.
    [Show abstract] [Hide abstract] ABSTRACT: Reaction-diffusion systems contribute to various morphogenetic processes, and can also be used as computation models in real and artificial chemistries. Evolving reaction-diffusion solutions automatically is interesting because it is otherwise difficult to engineer them to achieve a target pattern or to perform a desired task. However most of the existing work focuses on the optimization of parameters of a fixed reaction network. In this paper we extend this state of the art by also exploring the space of alternative reaction networks, with the help of GPU hardware. We compare parameter optimization and reaction network optimization on the evolution of reaction-diffusion solutions leading to simple spot patterns. Our results indicate that these two optimization modes tend to exhibit qualitatively different evolutionary dynamics: in the former, the fitness tends to improve continuously in gentle slopes, while the latter tends to exhibit large periods of stagnation followed by sudden jumps, a sign of punctuated equilibria.
    Full-text · Chapter · Oct 2011