Skip to Main content Skip to Navigation
Master thesis

Modèles de diffusion et estimation des dynamiques d'ADN tumoral circulant pour la détection d'une résistance à une thérapie ciblée

Abstract : This project is part of a project that aims to exploit circulating tumor DNA sequencing data (ctDNA). ctDNA is extracellular DNA from cancer cells released into the blood. These ctDNA observations are indirectly related to the sizes of the different cell populations in the tumor, including sensitive cells and cells resistant to targeted therapy. In this context, physicians wish, for instance, to use these observations to predict resistance or lack of resistance to a particular targeted therapy in patients with solid cancer. Our work is part of this modelling and estimation problem for cancer. We have implemented the appropriate tools and methods, considering the simplest case of dimension 1 tumor growth models, and to apply these probabilistic models to our real data, we must carefully define how to estimate the parameters. A classic method for parametric estimation of diffusion processes is the so-called ''maximum likelihood'' method based on continuous observations. These methods are based on advanced stochastic calculus results, such as Girsanov's theorem.
Complete list of metadata

Cited literature [53 references]  Display  Hide  Download

https://hal.inria.fr/hal-02194765
Contributor : Vincent Hass <>
Submitted on : Thursday, July 25, 2019 - 11:15:08 PM
Last modification on : Tuesday, May 18, 2021 - 2:14:03 PM

File

Stage_HASS_Vincent_HAL.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : hal-02194765, version 1

Collections

Citation

Vincent Hass. Modèles de diffusion et estimation des dynamiques d'ADN tumoral circulant pour la détection d'une résistance à une thérapie ciblée. Mathématiques [math]. 2018. ⟨hal-02194765⟩

Share

Metrics

Record views

165

Files downloads

543