Deep learning for drug analysis

• Main Author: Lim, Kayla Jia Yu
• Other Authors: Su Rong
• Format: Final Year Project
• Language: English
• Published: Nanyang Technological University 2024
• Subjects: Computer and Information Science; Engineering; Medicine, Health and Life Sciences; Drug-drug interaction; Deep-learning; Machine-learning
• Online Access: https://hdl.handle.net/10356/177185
• Institution: Nanyang Technological University

The increasing prevalence of drug-drug interactions (DDIs) poses a significant challenge to patient safety within the healthcare system. To address this issue, this project aims to enhance the predictive capabilities of DDI through the development and refinement of machine learning models. Leveraging extensive datasets and advanced methodologies, the project focuses on integrating Graph Neural Network (GNN) frameworks, used in Protein-Protein interaction (PPI) prediction, to improve DDI prediction accuracy. Despite encountering challenges such as technical issues and complex code debugging, the project explores new evaluation techniques inspired by GNN frameworks. While the debugging process did not yield a solution, valuable insights were gained, contributing to a deeper understanding of the code base, and debugging techniques. Moving forward, the project emphasizes the importance of continuous learning and refinement to address evolving challenges in DDI prediction and enhance patient care within the healthcare system.