A Multi-Dimensional Anomalous Space Modelling Approach to Understanding Viral Life Cycles

Project description

Viruses, though microscopic, exert profound impacts on individuals, societies, and economies. The COVID-19 pandemic exemplified this, with recent estimates published in the Medical Journal of Australia indicating that long COVID alone cost the nation approximately $10 billion in a single year. These figures underscore the urgent need for a deeper, data-driven understanding of viral behaviour and evolution. This research proposes a novel application of anomaly detection techniques to the study of viral genomics, specifically targeting the identification and prediction of viral mutation and cessation patterns. Traditional anomaly detection methods applied to virus life-cycle modelling often rely on a single observer or fixed reference point for decision-making, limiting their ability to capture the multi-dimensional complexity of viral evolution. To address this limitation, this project introduces a multi-dimensional anomalous space modelling framework that defines multiple observer points. Each observer point represents a distinct spectrum of normality, based on variable thresholds or biological rules derived from genomic, epidemiological, or environmental data. Instances that fail to conform to these defined normalities are identified as anomalies, potentially indicating viral mutations or adaptive responses. By analysing and comparing these anomalous spaces, the proposed approach aims to reveal the dynamic progression of viruses enabling the detection of mutations as they occur, or even before they manifest. Furthermore, the framework provides a mechanism to evaluate drug interactions and their potential role in inducing viral cessation. This interdisciplinary research bridges computer science, bioinformatics, and health science, fostering collaboration among data scientists, virologists, and epidemiologists. By integrating genomic, environmental, and epidemiological data streams, the project seeks to construct an evolving knowledge model capable of predicting viral trajectories. The ultimate objective is to support pre-emptive public health responses and contribute to global efforts to mitigate the impact of viral outbreaks. The successful implementation of this research will mark a step forward in computational virology, offering a dynamic and adaptive model for understanding and combating viral evolution demonstrating that there is indeed more to this tiny pathogen’s survival than meets the eye.

Co-supervisors

Dr Vijini Mallawaarachchi Flinders Accelerator for Microbiome Exploration College of Science & Engineering

Assumed knowledge

Programming in Python, AI/ML model training

Note: You need to register interest in projects from different supervisors (not a number of projects with the one supervisor).
You must also contact each supervisor directly to discuss both the project details and your suitability to undertake the project.