Non-Markovian Modeling of Gene Expression : Toward a More Realistic Description
Lundi 12 janvier 2015 14:00
- Duree : 1 heure
Lieu : Conference room - LIPhy - Bât E - 140 Avenue de la Physique - St Martin d’Hères. Accès par interphone, appeler le secrétariat
Orateur : Emanuele LEONCINI (Harvard Medical School, Boston, USA)
Gene expression is a fundamental process within prokaryotic and eukaryotic cells, and has been proven experimentally to be inherently stochastic. A key assumption in stochastic models of gene expression is that each step of the protein production process has an exponentially distributed duration, assuring thus a Markovian description of the system. Nevertheless, it is well-established that some steps, such as the protein elongation, do not show exponential durations.
The aim of this talk is to show that a description based on marked Poisson point processes (MPPP) is a natural modeling framework to tackle such difficulties. In particular, we illustrate the interest of this framework in gene expression modeling by providing analytical, closed formulas of the mean and variance of the copy-number of proteins, assuming non-exponential duration of protein elongation. A more appropriate deterministic description of protein dilution due to the cell growth is also considered and the resulting hybrid model is analyzed using the MPPP framework. The importance of the statistical assumptions in the modeling of protein production is shown also by a counter-intuitive result on the fluctuations of protein copy number when considering non-exponentially distributed durations.
Contact : hans.geiselmann@ujf-grenoble.fr
Discipline évènement : (Physique)
Entité organisatrice : (LIPhy)
Nature évènement : (Séminaire)
Evènement répétitif : (Séminaire Mat. Molle/grise)
Site de l'évènement : Domaine Universitaire de St Martin d’Hères
Prévenir un ami par email
Télécharger dans mon agenda