Documents disponibles écrits par cet auteur

Network Embedding for Classification Toward a new hybrid drug repurposing method / BACHRI Walid

Public ISBD Titre : Network Embedding for Classification Toward a new hybrid drug repurposing method Type de document : projet fin études Auteurs : BACHRI Walid, Auteur Langues : Français (fre) Catégories : BIG DATA Index. décimale : mast 269/19 Résumé : The abundance and the emergence of large scale biomedical and pharmacological data has open the doors to new opportunities for drug development, particularly for drug repurposing. In this project, we develop a new hybrid computational method combining heterogeneous network embedding and machine learning classification, to predict potential drug-target interactions which is known as drug repurposing. This method takes as input a heterogeneous network that consists of drugs and proteins related informations, then it focuses on learning the embedding of proteins and drugs, after that we construct an edge feature space containing a vector representation for each pair of drug and protein based on their own learned representations by using some binary operators, this edge feature space is used to train classifiers to predict the presence of an edge or its absence. This method achieves outstanding performance compared to the state of the art methods for computational drug repurposing. The predicted drug target interactions by our hybrid method need to be validated experimentally in a dedicated laboratory Network Embedding for Classification Toward a new hybrid drug repurposing method [projet fin études] / BACHRI Walid, Auteur . - [s.d.]. Langues : Français (fre) Catégories : BIG DATA Index. décimale : mast 269/19 Résumé : The abundance and the emergence of large scale biomedical and pharmacological data has open the doors to new opportunities for drug development, particularly for drug repurposing. In this project, we develop a new hybrid computational method combining heterogeneous network embedding and machine learning classification, to predict potential drug-target interactions which is known as drug repurposing. This method takes as input a heterogeneous network that consists of drugs and proteins related informations, then it focuses on learning the embedding of proteins and drugs, after that we construct an edge feature space containing a vector representation for each pair of drug and protein based on their own learned representations by using some binary operators, this edge feature space is used to train classifiers to predict the presence of an edge or its absence. This method achieves outstanding performance compared to the state of the art methods for computational drug repurposing. The predicted drug target interactions by our hybrid method need to be validated experimentally in a dedicated laboratory

Réservation

Réserver ce document

Exemplaires

Code barre	Cote	Support	Localisation	Section	Disponibilité
mast 269/19	mast 269/19 BAC	Texte imprimé	Unité des masters	Mast/19	Disponible