INP teams awarded MIC Cancer Inserm - ITMO Aviesan 2020 grant

Date de l'évènement: 
Mercredi, 22 Juillet, 2020

We are very excited to announce that our project "Artificial Intelligence for Onco-Plasma" carried out by Remi Eyraud from QARMA team of LIS in collaboration with team 8team 9, team 10and PINT platform of INP, was awarded a grant from the MIC (Mathemathiques et Informatique) Cancer Inserm – ITMO Aviesan 2020 call "Approches interdisciplinaires des processus oncogéniques et perspectives thérapeutiques : Apports à l’oncologie des mathématiques et de l’informatique" 

Abstract : The objective of this project is to create a universal and non-invasive method to help diagnose cancers. The proposed method is based on the existence of unique plasma denaturation profiles (signatures) for different cancers. The plasma denaturation profile represents the total denaturation curve (under the influence of temperature) of its constituent proteins. Due to homeostasis, the plasma denaturation profiles of healthy individuals do not vary significantly. However,in the presence of a  disease, the composition of the blood or the thermal stability of circulating proteins may change, thus altering the plasma denaturation profile. Three different instruments can be used to obtain these denaturation profiles: the PEAQ-DSC instrument (our reference) and two nano-DSF, namely the research oriented Prometheus NT.Plex and the new Tycho. Our goal is to study the numerical outputs of these three instruments on plasma samples from 4 cohorts of blood plasma from patients with glioma, melanoma, lung or colorectal cancer, as well as from healthy donors (oncologists we are working closely with promised at least 100 samples in each cohort).

The core of this proposal is to study these data using modern artificial intelligence approaches, that is, machine learning methods. This project aims to follow four complementary lines of research:
(1)  Adaptation of classical algorithms to handle the particularities of the data
(2)  Development of multi-view learning algorithms, a recent field whose framework fits well these data
(3)  Design new artificial neural network architectures dedicated to the context of this project, for scalability
(4)  Interpretability and explainability of the behavior of the obtained models

=> Stay tuned for the announcements of a 2-year position for a computer science engineer and 3-year position for a biophysics engineer (IGE) <=

Contact : 

See full 2020 Plan Cancer results :