par Haye, Alexandre 
;Albert, Jaroslav ;Rooman, Marianne
Référence Lecture notes in computer science, 6685 LNBI, page (242-252)
Publication Publié, 2011
;Albert, Jaroslav ;Rooman, Marianne Référence Lecture notes in computer science, 6685 LNBI, page (242-252)
Publication Publié, 2011
Article révisé par les pairs
| Résumé : | The evolution of the gene expression levels of Drosophila melanogaster, from the embryonic to adult development phases, has been studied on the basis of a microarray time series involving the expression levels of more than 4000 genes over 67 time-points, and has been modeled by a system of linear differential equations with constant coefficients. Here we investigate the robustness of this model against perturbations of its parameters and of the initial data values. We found that the model is not robust at all for fully connected networks, but that the robustness significantly increases after parameter reduction. This puts some limits to the biological relevance of linear models for gene expression evolution. © 2011 Springer-Verlag Berlin Heidelberg. |
