Séminaire Open MAMBA

Présentation:

Le séminaire Open MAMBA est un séminaire interne mais ouvert à tous, ayant lieu une fois par mois sur le site INRIA Paris. Les présentations sont effectuées par les étudiants en thèse/post-doctorant de l’équipe MAMBA et le séminaire a pour buts:

  •  Renforcer l’esprit d’équipe  en centrant les présentations sur les jeunes membres de l’équipe MAMBA (étendue)
  •  Apporter plus de math-bio au centre INRIA (en complément des rencontres INRIA-LJLL)
  • Favoriser les échanges scientifiques sur les projets en cours, chaque présentation étant suivie par une discussion informelle par les chercheurs.

Le séminaire aura lieu une fois par mois les lunid de 13h30 à 15h au centre INRIA Paris. Il est composé de 2 présentations de 30 minutes chacune suivies de discussions de 15mn.

Localisation:

INRIA Paris, 2 rue Simone Iff (voir Comment s’y rendre), Salle ‹‹ Jacques-Louis Lions ››, bâtiment C

NB: Une carte d’identité sera demandée à l’entrée du batiment

Organisateurs:

Diane Peurichard: diane.a.peurichard@inria.fr

Dirk Drasdo: dirk.drasdo@inria.fr

Présentations:

 

Lundi 17 juin 2019, 13h30, salle de séminaire du L. J-L. Lions

      1. Mathieu Mezache

Title: Testing for high-frequency features in a noisy signal: applications to prion aggregation process experiments

Abstract: We present a procedure which detects high frequency (HF) features in a noisy signal. We propose a parametric characterization on the Fourier domain of the HF features. Then we introduce a procedure to evaluate these parameters and compute a p-value which assesses in a quantitative manner the presence or absence of such features, that we also call ”oscillations”. The procedure is well adapted for real 1-dimensional signals. If the signal analyzed has singular events in the low frequencies, the first step is a data-driven regularization of its Fourier transform. In the second step, the HF features parameters are estimated. The third step is the computation of the p-value thanks to a Monte Carlo procedure. The test is conducted on signals from Prion disease experiments.

2. Hugo Martin

Title: Hétérogénéité des durées des durées des cycles cellulaires de levures en l’absence de télomérase : analyse de données et modélisation

Abstract:

Aux extrémités de chaque chromosome se trouve une région hautement répétitive et non codante appelée télomère. Au cours des divisions de la cellule, ces télomères sont soumis à (au moins) deux mécanismes : un de raccourcissement, au fur et à mesure que les générations se suivent, et un autre d’élongation, par l’action d’une holoenzyme appelée télomérase.   Afin d’étudier le phénomène de vieillissement cellulaire, cet enzyme est désactivée, et un certain nombre lignées suivies individuellement. Ainsi, on obtient la durée de chaque cycle cellulaire et leur succession dans les lignées, jusqu’à éventuellement atteindre un état dit de sénescence réplicative, dans lequel les organismes étudiés (des levures) ne se dévident plus ou peu. Le modèle biologique standard prévoit qu’après désactivation de la télomérase, les levures vivent un certains nombre de cycles de longueur « normale » (~90 minutes) avant d’atteindre la phase de sénescence, puis de mourir après quelques générations. Cependant, on constate expérimentalement que des cycles anormalement longs se produisent avant l’entrée en sénescence, et sont suivis par des cycles de durée attendue, et ce en nombre trop grand pour que l’hypothèse de l’hétérogénéité propre à la biologie soit satisfaisante. On essaie alors de comprendre la présence de ces « anomalies » (par rapport au modèle biologique).
Je commencerai par présenter rapidement les résultats obtenus par nos collègues
biologistes et ceux obtenus lors d’une précédente collaboration entre nos deux équipes, puis je parlerai de mes résultats, qui consistent en une réanalyse des données expérimentales et un peu de modélisation. 

Lundi 11 mars 2019, 13h30, salle de séminaire du L. J.L. Lions:

 

Cecile Della Valle

Titre: Inverse problem and regularization method applied to protein dynamics 

 
Abstract: In this talk, I am interested in the evolution of the polymer size distribution modeled by the polymerization-depolymerization dynamics ( the Lifschitz-Slyosov PDE system). Our inverse problem consists in reconstructing the initial distribution of polymers knowing the parameters of the chemical reactions and measurements on the studied sample. To address this non-linear problem, we first rely on the study of the direct problem in order to reduce ourselves to solving a transport equation. I will then present the two methods that we are developing: a regularization method based on a method by B. Perthame and J. Zubelli in 2006, then a data assimilation approach (Kalman filter).
 

Julia Delacour

Titre: A model for the formation of p_{62}-Ubiquitin aggregates involved in cellular autophagy 

Abstract: Cellular autophagy is a key pathway that allows the cell to degrade nocive or dysfunctional elements (such as old mitochondriae), and that has been involved in a wide range of disease from neurodegenerative to cancer. In this talk, we will introduce and study a model for p_{62}-Ubiquitin aggregates, that have been shown (Zaffagnini et al., 2018) to play a key-role in cellular autophagy. We will begin by introducing the model for the formation of one p_{62}-Ubiquitin aggregate. We will exhibit three different regimes of this model, supported by numerical simulations that will be shown. We will then present the mathematical analysis of each regime that requires some dynamical systems theory tools. Finally, if time is left, we will briefly introduce the equation for the formation of several p_{62}-Ubiquitin aggregates, that is just a transport-coagulation equation directly derived from the previous model. 

Lundi 21 janvier 2019


 

Laura Kanzler

 
Title: KINETIC MODELLING OF COLONIES OF MYXOBACTERIA
Abstract: Myxobacteria are rod-shaped, social bacteria that are able to move on flat surfaces by « gliding » and form a fascinating example of how simple cell-cell interaction rules can lead to emergent, collective behaviour. Observed movement patterns of individual bacteria include straight runs with approximately constant velocity, alignment interactions and velocity reversals.
The fact that this behaviour is a consequence of a direct cell-cell contact interaction rather than by diffusing chemical signals indicates the suitability of kinetic modelling.
In this talk a new kinetic model of Boltzmann-type for colonies of myxobacteria will be derived and investigated. An existence and uniqueness result will be shown in a special setting for the spatial homogeneous case, as well as exponential decay to an equilibrium.
The talk will be concluded with numerical simulations.

Lundi 19 mars 2018

 

1.Hugo Martin

Titre: Measure solution for the classical size-equation with linear growth rate

Abstract: In a recent paper, E. Bernard, M. Doumic and P. Gabriel showed the existence of a cyclic asymptotic bahaviour of the solution of the size-equation with linear growth rate in a weighted L^2 space which consists in a Fourier serie built on the eigenfunctions of the model. On another hand, P. Gabriel recently solved the age Equation and showed a relation with the related principal eigenfunction. We are now wondering how this behaviour appear for measure solutions in the case of the size Equation. After defining what a measure solution to a PDE is, I will explain the semigroups methods we attempt to use to obtain the desired result and the difficulties we are already facing.

2. Camille Pouchol

Titre: Dirichlet control towards steady states for the 1D monostable and bistable equations

Abstract: I will talk about the problem of controlling parabolic semilinear equations arising in population dynamics, namely the 1D monostable and bistable equations on (0,L) for a density of individuals between 0 (extinction) and 1 (invasion), by means of Dirichlet controls between 0 and 1. I will explain why driving the system towards 0 or 1 asymptotically is always possible for the state 1, and for 0 as well if L is below some threshold value, computable by phase plane analysis. In the bistable case, I will prove that the other homogeneous steady state 0 < θ < 1, though unstable for the corresponding ODE, can be reached in finite time. The central tools are the phase portrait for the stationary problem, the parabolic maximum principle, and the staircase method in control theory.

Lundi 12 février 2018

1. Mathieu Mezache:

Title:An oscillatory kinetic model for the Prion aggregation process.
Abstract: We investigate the oscillatory behaviour of the PrP protein during the polymerization/depolymerization process. In order to modelize this oscillatory process, we study a simplified Belousov Zhabotinsky reaction from a kinetic point of view. This simplified oscillatory system of chemical reactions allows us to introduce a modified Becker-Döring system where the trajectories oscillate. A key to have a closed oscillatory polymerization/depolymerization system is to consider different specices of polymers and monomers. The main goal is to investigate the size distribution in the polymerisation /depolymerisation process of the PrP protein. »

2. Martin Strugarek

Title: Mosquito population replacement: an optimal control viewpoint
(in collaboration with Luis Almeida, Yannick Privat (LJLL) and Nicolas Vauchelet (LAGA))


Abstract: Transinfection of Wolbachia bacteria is currently used in wild Aedes populations to suppress these mosquitoes’ ability to transmit some viruses (vector competence), in particular dengue viruses. This recently developed technique relies on the mass rearing and release of lab-infected individuals of both sexes. The introduced phenotype can replace the wild one thanks to an asymmetric sterile crossing known as cytoplasmic incompatibility (CI). It is expected to practically stop disease circulation and be very cost-efficient as the Wolbachia infection will maintain once established over many generations, thanks to almost perfect vertical transmission from mother to offspring. Our work aims at improving these so-called « Wolbachia population replacement strategies », using mathematical modeling to design the release sizes and timings (and, in future, locations). Here, we formulate this as an optimal control problem for an ODE system and prove qualitative properties of the optimal strategies under resource constraints. Numerical simulations illustrate these results. We also recover a bistable scalar equation on the proportion of infected individuals by a proper scaling of the parameters, and rigorously prove the convergence of the optimization problems to a limit problem for which uniqueness and explicit solution are established.

Lundi 22 janvier 2018

1. Yi Yin:

2. Jieling Zhao:

Title: A dynamic activator-inhibitor system to model liver fibrosis at early stage

Abstract: Liver fibrosis is a characteristic of most types of liver disease due to chronic liver damage. During the process of fibrosis, excessive accumulation of extracellular matrix (ECM) proteins, such as type I and III collagen are produced by activated fibroblast-type cells in the liver, which leads to cirrhosis and eventually the failure of the liver. Due to the complexity of the whole process, the mechanism behind is still elusive. Computational modeling aspects of the liver fibrosis may provide some insights towards the understanding of the mechanism and developing novel treatment strategies. In this study, we develop a dynamic activator-inhibitor system to study the liver fibrosis at early stage. Our modeled pattern of the activator which promotes the ECM deposition can roughly capture the observed ECM pattern upon chronic liver injury. This result can be used as preliminary pattern as guidance for the further exploration of the formation of ECM pattern observed in vivo.