Description
Date depot: 3 juin 2021
Titre: Deep learning to predict protein-protein interaction networks for environmental microbial communities
Directrice de thèse:
Alessandra CARBONE (LCQB)
Directrice de thèse:
Lucie BITTNER (ISYEB: UMR 7205)
Domaine scientifique: Sciences et technologies de l'information et de la communication
Thématique CNRS : Sciences de l’information et sciences du vivant
Resumé: Metagenomics provides a huge inventory of species present in environments and of metabolic functions performed by environmental communities. It offers enormous potential for discoveries, as more than 99% of microbial species cannot be cultivated in the laboratory. It has given rise to several large-scale projects to characterize the microbial diversity of the oceans (e.g., GOS, Tara Oceans, Malaspina, OSD), of microbes in symbiosis with humans, of the composition of soils, urban environments, or subjected to extreme conditions (eXtreme Microbiome Project). Each metagenomics experiment generates large amounts of raw data (on the order of several terabytes of sequence per sample), the processing of which presents several algorithmic and data analysis / learning challenges.
The goal of this PhD project will be to develop new computational approaches based on deep learning to reconstruct protein-protein interaction (PPI) networks for metagenomic samples starting from sequence reads. The aim is to predict PPI networks that allow a community of microbes to perform their metabolic functions. Questions on biogeography and evolution of PPIs will be addressed with a comparison of PPI though samples from different ecosystems.
Doctorant.e: Volzhenin Konstantin