Loading...

Drug-Target Interaction Prediction with Deep Learning and Recommender Systems

Nosrati, Amir Hossein | 2021

153 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: English
  3. Document No: 55348 (52)
  4. University: Sharif University of Technology, International Campus, Kish Island
  5. Department: Science and Engineering
  6. Advisor(s): Ghafourian Ghahramani, Amir Ali; Kavousi, Kaveh
  7. Abstract:
  8. A drug can be defined as a substance made to prevent disease, cure a specific symptom, relieve pain, and reduce anomalies in the body. The process of drug designing is so laborious, complex, costly, and time-consuming that chance of failure during the lab experiment stages is high. These challenges have persuaded researchers to find new usage for existing drugs, referred to as drug repurposing, with the main advantage of reducing cost, risk, and time. To this aim, computational methods have been applied to discover hidden pharmaceutical capabilities of drugs in terms of predicting whether a particular drug can interact with a particular protein.Graph Neural Networks (GNNs) have recently gained significant accomplishments in supervised learning on graphs by integrating node information and topological structure. Thus, GNNs can significantly advance the performance of recommender systems. This thesis focuses on drug-target prediction, and we represent the interactions between drugs and target proteins through bipartite networks. In addition, we create a network between proteins based on their structural similarities. The task of drug-target prediction can be formulated as the link prediction in a bipartite network. To be more precise, we propose a GNN framework to predict the interaction between drugs and proteins. The proposed GNN framework tries to learn the latent factors of drugs and proteins by jointly utilizing the information provided by two networks: the bipartite drug-protein network and the protein similarity network.To evaluate the performance of the proposed GNN framework, we prepared a dataset containing relevant information about drugs, target proteins, and their interactions. We also used a famous benchmark dataset containing interactions between drugs and different groups of enzymes, ion channels, GPCRs, and nuclear receptors. The results indicate that the proposed framework exhibits acceptable performance and can get better results compared to some proposed methods in the literature. On the first dataset, our method achieved 92% and 85% of accuracy over training and test sets, respectively. For the second dataset, accuracies are 89%, 86%, 82%, and 80%, respectively, on four important classes of targets
  9. Keywords:
  10. Graph Neural Network ; Drug-Target Interaction ; Bi-Partite Drug-Protein Network ; Deep Learning ; Recommender System ; Drug Repurposing

 Digital Object List

 Bookmark

No TOC