Skip to content. | Skip to navigation

Personal tools
Log in
Sections
You are here: Home Research Research teams Numed - Integrated models for biomedical research

Numed - Integrated models for biomedical research

Disciplines Mathematics, Medicine - Health
Research fields Neurobiology, Cancerology, Infectiology, Multiscale models, Data management
Supporting organisms CNRS, ENS de Lyon, Inria
Geographical location Grenoble
Lab Inria
Team leader Emmanuel Grenier
Webpage https://www.inria.fr/equipes/numed

 

We are working within 3 applied areas (neurological diseases, cancer diseases and infection diseases) and 3 methodological axes (parameter estimation, knowledge management and multiscale approach for modeling).

Neurological diseases
Our aim is to model the main pathophysiological mechanisms involved in an ischemic stroke and to use this model to carry out in silico experiments in order to better understand the interactions of these mechanisms and their influences on cell damage and to explore the effects of various therapeutic strategies in stroke patients.
The global model of stroke is composed of eight sub-models which represent the following phenomena: metabolic reactions, ionic movements, glutamate excitotoxicity, cytotoxic oedema development, spreading depressions, inflammatory reaction, free radical synthesis and cell death by necrosis or apoptosis. These sub-models were built independently and are currently under study.

Cancer diseases
Our cancer modeling projects mostly focus on the optimization of antiangiogenic drugs delivery. A current project involving statisticians, pharmacologists, cliniciens and experimentalists is aimed at optimizing the delivery of antiangiogenic therapy Sunitinib (Sutent, Pfizer Inc) in combination with chemotherapy 5FU and Irinotecan in xenografted mice. Treatment effect together with the dynamic of tumor growth is being analyzed in more than one hundred mice.
Theoretical models, focusing on the interplay between the dynamic of tumor growth and the dynamic of the angiogenesis process are also developed. Within a multiscale approach, our team is building a complex model of antiangiogenic drug effect bringing together molecular description of the VEGF-VEGFR2 pathway with macroscopic simulation of tumor growth and blood vessel formation. The same approach is also applied to the MAPK pathway in concert with biologists from the group of Yossi Yarden at the Weizmann Institute of Science.
The team is involved in the ETOILE project with the aim to help thanks to computational methods for the prediction of hadrontherapy efficacy and for a rational selection of responders. In the ETOILE framework, the team is involved in the European project ULICE - Union of Light-Ion Centres in Europe.

Infection diseases
In collaboration with MERIAL a health animal company, we develop mechanistic models to predict the dynamic of Feline immunodeficiency virus (FIV) primary infection analysing multiple biomarkers of the immune system. We aim at developping a faithful model to optimize the development of a vaccination strategy.
In collaboration with the department of public health in Edouard Herriot Hospital in Lyon (Prof. Phillippe Vanhems), based on the CASCADE database, we explore the dynamic of primary infection in humans.

Parameter estimations
We use a population integrating different levels of variability in our mixed-effect models. In such a context, model parameters are defined by a fixed (mean) effect and a radom (inter-individual) effect. We use the MONOLIX software for parameter estimation.

Multiscale modeling
The multiscale modeling approach consists of building model with the facility to integrate data coming from different levels of the organism. For istance, molecular biomarkers of a given pathological process, macroscopic data such as tumor size, clinical scrores and survival. Four years ago, we have started the development of a complex mechanistic model of tumor growth. We focused this model on the subcellular level (molecular and genetic network), the cell cycle, and the tissue scale by means of partial differential equations coupled to discrete formalisms.

Knowledge management
Knwoledge management is a crucial issue for modeling. We are involved in the development of a quantitative database to store parameters values from the literature in order to support mechanistic modeling of complex biological processes.