Passer au contenu principal
Publiez votre CV - Laissez les employeurs vous trouver

emplois univ. evry paris saclay

Trier par : -
3 offres d'emploi

Job Post Details

stage M2 - Ingénieur /Deep Learning appliqué au domaine de la santé - job post

Univ. Evry - Paris Saclay
91000 ÉvryTélétravail partiel
Temps plein, Stage

Détails de l'offre

Correspondance entre ce poste et votre profil.

Type de poste

  • Temps plein
  • Stage

Avantages
Extraits de la description complète du poste

  • Prise en charge du transport quotidien

Description du poste

Proposition de stage de recherche de niveau Master 2 ou Ingénieur

Titre : Graph Neural Networks for Sepsis Prediction from multi-omics Data

Mots-clefs : Graph neural networks, bioinformatics, Sepsis prediction

Encadrants : Farida Zehraoui, Blaise Hanczar et Victoria Bourgeais

Laboratoire d’accueil et lieu de stage : Laboratoire IBISC-IBGBI. Univ. Evry, université Paris-Saclay. 23 boulevard de France. 91034 Evry

Descriptif :

Precision medicine, which is also known as stratified medicine, uses the genomic characteristics of patients to provide personalized care. Thanks to "omics" technologies, such as genomics (DNA sequencing), transcriptomics (microarrays), and proteomics, we can now generate massive quantities of genomic data on patients. These data can cover all the mechanisms involved in the variations that occur in the cellular networks that influence the functioning of organ systems in humans. They can be used for diagnosis, prognosis, prediction of personalized patient treatment, and more. Machine learning, especially Deep Learning, has become a promising tool over the past decade to support precision medicine.

In the context of the IHU Project PROMETHEUS about early Sepsis prediction, the aim of this internship is to develop a graph neural network (GNN) based approach that uses multiple data sources representing different types of omics associated with patients.

In addition to the multi-omics integration, we will constrain the model by domain knowledge. The integration of biological domain knowledge, such as molecular interaction networks and biological ontologies, can incorporate a good inductive bias that can improve the generalization of the model. This can also improve the interpretability of the model by ensuring that the learned high-level features are consistent with domain knowledge.

Bibliography

- V. Bourgeais, F. Zehraoui, B. Hanczar. GraphGONet: a self-explaining graph-based neural network encapsulating the Gene Ontology for phenotype prediction on gene expression. Bioinformatics, 2022.

- V. Bourgeais, F. Zehraroui, B. Hanczar. Deep GONet: Self-explainable deep neural network based on Gene Ontology for phenotype prediction from gene expression data. BMC bioinformatics 22, 455, 2021.

- Graw, S., Chappell, K., Washam, C.L., Gies, A.J., Bird, J.T., Robeson, M.S., & Byrum, S.D. (2020). Multi-omics data integration considerations and study design for biological systems and disease. Molecular omics.

- B. Hanczar, F. Zehraoui, T. Issa, M. Arles. Biological interpretation of deep neural network for phenotype prediction based on gene expression. BMC bioinformatics 21 (1), 1-18, 2020.

Type d'emploi : Temps plein, Stage
Durée du contrat : 6 mois

Rémunération : 4,35€ par heure

Nombre d'heures : 35 par semaine

Avantages :

  • Prise en charge du transport quotidien

Exigences linguistiques flexibles :

  • Français non requis

Programmation :

  • Du lundi au vendredi
  • Travail en journée

Lieu du poste : Télétravail hybride (91000 Évry)

Date limite de candidature : 31/12/2021
Date de début prévue : 03/01/2022

Permettez aux employeurs de vous trouver.Téléchargez votre CV