Skip to Main content Skip to Navigation
Theses

Méthodes computationnelles pour améliorer les phases primaires de recherche de nouveaux médicaments

Abstract : The drug discovery process is long, costly and very risky. The objective of this doctoral thesis is to improve the relevance of the primary phases of drug discovery by developing computational methods. The first contribution concerns the development of the Pegasus knowledge graph in order to capitalize on the heterogeneous and multi-sourced pharmaco-biological data of the pharmaceutical sector. The industrial applications of Pegasus address the problems of therapeutic projects and allow to characterize off-target effects of perturbators, to design a new experiment, and to identify focused screening libraries. The second contribution concerns the development of an algorithm for the identification of positive control compounds and a normalization algorithm to improve the design and analysis of high content phenotypic screening experiments. These algorithms allow the normalization of phenotypic signatures obtained from screening campaigns and the integration of informative phenotypic similarities in the Pegasus knowledge graph. The third contribution concerns the development of a mathematical model of the microtubule tyrosination cycle which explains, on the one hand, the inactivity of chemical compounds shown to be active outside the cell, and on the other hand, suggests the need to activate two reactions of this cycle, in synergy, to obtain an effect in cellular models. This illustrates the contribution of mathematical modeling to, on the one hand, predict and understand the counter-intuitive dynamics of biochemical processes that are not representable by static knowledge graphs as in Pegasus, and on the other hand, guide the design of new screening experiments. The scientific contributions and industrial applications of this thesis are developed in the context of the primary phases of drug discovery and are intended to extend to the clinical phases of the pharmaceutical process.
Complete list of metadata

https://tel.archives-ouvertes.fr/tel-03715715
Contributor : ABES STAR :  Contact
Submitted on : Wednesday, July 6, 2022 - 4:44:14 PM
Last modification on : Monday, September 5, 2022 - 3:49:21 PM

File

108640_GRIGNARD_2022_archivage...
Version validated by the jury (STAR)

Identifiers

  • HAL Id : tel-03715715, version 1

Citation

Jeremy Grignard. Méthodes computationnelles pour améliorer les phases primaires de recherche de nouveaux médicaments. Intelligence artificielle [cs.AI]. Institut Polytechnique de Paris, 2022. Français. ⟨NNT : 2022IPPAX045⟩. ⟨tel-03715715⟩

Share

Metrics

Record views

100

Files downloads

44