An overview of graph neural networks for molecular biology: Challenges and solutions

Divyansh Choubisa *

Department of Computer Science and Engineering, Wilfrid Laurier University, Ontario, Canada.
 
Review
International Journal of Science and Research Archive, 2024, 13(01), 2670–2673
Article DOI: 10.30574/ijsra.2024.13.1.1985
Publication history: 
Received on 08 September 2024; revised on 04 October 2024; accepted on 17 October 2024
 
Abstract: 
This review explores the application of graph neural networks (GNNs) in molecular biology, with a particular focus on their role in drug discovery. Various use cases, including molecule classification for cardiotoxicity detection and the prediction of molecular properties have been examined. A comprehensive analysis compares methodologies, problem statements, datasets, and findings across different studies. A core themes of review is the intuition to regularize a graph neural network which was proven to be reliable in terms of robustness for drug discovery purposes using noise nodes techniques. In the conclusion section, key insights from the reviewed literature and promising future directions for research have been presented. This work aims to provide a foundational understanding and guide future innovations in leveraging GNNs for molecular applications.
 
Keywords: 
Graph Neural Networks; Drug Discovery; AI in Medicine; Healthcare AI; Computational Biology
 
Full text article in PDF: 
 
Copyright information: 

Copyright © 2024 Author(s) retain the copyright of this article. This article is published under the terms of the Creative Commons Attribution Liscense 4.0