Description du projet
Une nouvelle manière d’aborder la résistance aux médicaments anticancéreux
Les avancées en génétique du cancer ont permis la découverte de nombreux traitements anticancéreux prometteurs et ciblés. Toutefois, la médecine de précision du cancer est souvent freinée par l’émergence de la résistance aux médicaments. Les scientifiques du projet COMBAT-RES, financé par l’UE, entendent répondre à ce problème en développant des méthodes d’identification de la résistance aux médicaments, de détection des biomarqueurs associés, et de prédiction ultérieure des combinaisons de médicaments visant à surmonter la résistance inévitable aux médicaments découlant de la monothérapie. Pour y parvenir, ils développeront des méthodes informatiques innovantes pour des criblages à haut débit de médicaments in vitro, et valideront ces prédictions in vivo. L’objectif à long terme est d’anticiper l’évolution du cancer et de prévenir la résistance à la monothérapie en identifiant des combinaisons intelligentes de médicaments qui présentent une action synergique et améliorent l’index thérapeutique.
Objectif
Personalising treatments based on tumour genetic profiles enables cancer precision medicine. However, treating cancers using targeted therapies often fails due to the emergence of drug resistance. Here, my goal is to use drug high-throughput screens (HTS) combined with computational methods to identify resistance and its biomarkers, and to overcome it with smart drug combinations to empower cancer precision medicine.
Identifying resistance in HTS is challenging: dissecting meaningful drug responses at high concentrations is impossible due to cytotoxicity, making non-responders and resistant cell lines indistinguishable, thus limiting resistance biomarker discovery to frequently mutated cancer genes. To address this, I will employ three approaches: 1) systematically identify non-responding cell lines carrying low-frequency resistance markers; 2) reveal intrinsic resistance driven by gene expression plasticity by conducting my own RNA sequencing experiments and modelling the maximal effect at high drug concentration; 3) identify drugs which increase cell viability, combined with drugs targeting fast proliferating cells. My paradigm shift, that resistance biomarkers become synergy markers, empowers smart drug combinations.
Additionally, I aim to predict drug synergy based on multi-task deep learning using molecular characterisation, QSAR modelling and monotherapies; and, to boost biomarker discovery by identifying clinically-relevant cancer subtypes based on transfer and reinforcement learning.
COMBAT-RES will benefit from data access to a phase III clinical trial in colorectal cancer (COREAD) and access to the largest human pancreas adenocarcinoma (PAAD) combination HTS (currently unpublished) accelerating the delivery of medicine for COREAD and PAAD patients. COMBAT-RES will interrogate the underpinnings of drug resistance, clinically-relevant subtypes and overcome it with highly synergistic drug combinations, enabling the next generation of precision medicine.
Champ scientifique
- natural sciencescomputer and information sciencesartificial intelligencemachine learningreinforcement learning
- medical and health sciencesbasic medicinepharmacology and pharmacydrug resistance
- medical and health sciencesclinical medicineoncologycolorectal cancer
- medical and health scienceshealth sciencespersonalized medicine
- natural sciencescomputer and information sciencesartificial intelligencemachine learningdeep learning
Mots‑clés
Programme(s)
Thème(s)
Régime de financement
ERC-STG - Starting GrantInstitution d’accueil
85764 Neuherberg
Allemagne