A. Prof. Dr. Binh Nguyen, Victoria University of Wellington, New Zealand

Biography: Dr Binh Nguyen has strong experience in data science and predictive modelling using deep learning, especially in the highly complex interdisciplinary research developed working on both academic and industry focused projects. In the past, he has won several awards for his work on the development of innovative medical devices and the application of data science to improve healthcare. Since joining Victoria University of Wellington (New Zealand - NZ) as a Senior Lecturer in Data Science in 2018, he has been focusing on developing novel methods based on deep learning and graph machine learning to solve emerging problems in health informatics, bioinformatics and drug discovery. He is currently the principal investigator of a prestigious project funded by Callaghan Innovation (NZ), aimed at using deep learning approaches to develop new natural products for treatment of type II diabetes. He is also the sole principal investigator of a research project in collaboration with the Singapore Immunology Network to develop a novel deep learning framework to identify FDA-approved drugs that can be repurposed for treatment of tuberculosis. In addition, he is a key researcher on other four projects funded by the Ministry of Business, Innovation and Employment (NZ) and the NZ Health Research Council with a total amount of more than NZ$18 million.

Research Areas: 

① Data science
② Machine Learning and Deep Learning
③ Health Informatics
④ Bioinformatics

Speech Title: Graph machine learning in drug discovery and development

Abstract: Contributing immeasurably to human health, drug development is a complex, expensive, and time-consuming endeavour. To improve time- and cost-efficiency, machine learning is increasingly being used in different stages of drug development, especially in drug discovery – the first stage to discover new drug candidates. However, traditional machine learning has not performed well in some specific drug discovery and development tasks due to lacking labelled data and losing important target structure information after data transformation. An increasingly popular technique is graph machine learning where complex relationships in data can be modelled. Due to the fact that biomedical data are naturally interconnected, a number of methods using graph machine learning have been proposed for drug discovery and development with great success. Those state-of-the-art methods and their applications will be discussed in this talk.