UQ Centre for Clinical Research Winter Research projects
A machine learning-based investigation into adverse events and vaccine-vaccine interactions for Japanese Encephalitis vaccines in Australia
Hours of engagement | 36 |
Location | Herston: UQ Centre for Clinical Research (918) |
Project description | With the help of machine learning methods, this project concentrates on the safety of Japanese encephalitis (JE) vaccination. In travel medicine, individuals heading to JE-endemic areas often receive JE vaccine, often along with other travel-related vaccines (e.g., rabies). However, limited evidence exists regarding how different JE vaccine formulations either alone or in combination with other vaccines may influence the risk of adverse events following immunisation (AEFIs). Project Aims and Approach: The student will utilize a large dataset from an active vaccine safety surveillance system in Australia, focusing on both the chimeric live attenuated and the inactivated JE vaccines. Supervised and unsupervised machine learning algorithms will be applied to detect patterns and identify specific combinations of JE vaccines (and co-administered vaccines) that pose elevated AEFI risks. Student will gain hands-on experience in data cleaning, feature engineering, model training, and validation using Python and related machine learning libraries. Finally, the machine learning outputs will be translated into actionable insights, aiming to optimize vaccine administration schedules and mitigate adverse events in travel medicine settings. |
Expected outcomes and deliverables | The student will a gain insight of machine learning techniques, hands on data analysis, and contribute to the preparation of a scientific paper. The deliverables of this project include the models developed in machine learning tools, a software program in Python if required, a publication, and conference presentations. |
Suitable for | The project is suitable for students with basic data analysis knowledge and understanding. Programming skills and experience in Python are desirable but not required. |
Primary Supervisor | |
Instructions to applicants | The supervisor MUST be contacted by students prior to submission of an application. |
Advancing spatial epidemiology with machine learning: A systematic review of methodological comparisons
Hours of engagement | 36 |
Location | Herston: UQ Centre for Clinical Research (918) |
Project description | Machine learning (ML) has revolutionized spatial epidemiology by enhancing disease modelling and prediction. However, its comparative performance against traditional spatial epidemiological models remains underexplored. This four-week Winter Research School project provides students with the opportunity to contribute to a systematic review that evaluates ML's role in advancing spatial epidemiology for infectious disease modelling. Participants will assist in refining the search strategy, screening studies, and extracting relevant data from peer-reviewed literature. The review follows PRISMA guidelines and focuses on studies that explicitly compare ML methods with traditional spatial models, including ensemble/hybrid approaches. Key research questions include: Which ML techniques have been used to improve traditional spatial epidemiology models? |
Expected outcomes and deliverables | By the end of this four-week Winter Research School project, participants will have contributed to key stages of a systematic review on machine learning (ML) applications in spatial epidemiology. The expected outcomes and deliverables include: Refined Search Strategy: Participants will help finalize the systematic search strategy across databases (Scopus, PubMed, Web of Science, Embase) to ensure comprehensive coverage of relevant studies. Screening of Studies: Using tools like Covidence or Rayyan, participants will assist in title and abstract screening based on predefined inclusion and exclusion criteria. Data Extraction Framework: Students will help develop a standardized data extraction form to collect key information on ML models, spatial methods, study design, and performance metrics. Preliminary Analysis: A summary of key trends in the reviewed studies, including commonly used ML techniques, their comparative performance, and methodological strengths/limitations. Research Report & Presentation: Participants will prepare a short research summary outlining their contributions, findings from the initial stages of the review, and potential next steps. A final presentation may also be given to summarize key insights. |
Suitable for | This project is ideal for students with an interest in: Public Health & Epidemiology: Particularly those interested in infectious disease modeling and spatial epidemiology. Data Science & Machine Learning: Students keen on exploring how ML techniques are applied to health research. Geospatial Analysis: Those with a background in GIS, spatial statistics, or environmental health. No prior experience in machine learning is required, but familiarity with research databases (e.g., PubMed, Scopus) and basic data handling would be beneficial. The project is well-suited for undergraduate (final-year) and postgraduate students looking to enhance their research skills. |
Primary Supervisor | Dr Behzad Kiani |
Instructions to applicants | The supervisor CAN be contacted by students prior to submission of an application. |
Dementia in Parkinson's Disease: An implementation roadmap
Hours of engagement | 25 |
Location | Herston: UQ Centre for Clinical Research (918) |
Project description | Parkinson's Disease dementia is the second most common type of dementia. Over 80% of people living with Parkinson's Disease develop dementia. Identification of an impending dementia is difficult as the clinical presentation of symptoms is unique. Best practice guidelines for cognitive evaluation of people living in PD have been developed to solve this problem. To complement the guidelines, a customised, scalable, and technologically advanced platform, called PDCogniCare, will manage routine cognitive assessments in people living with PD. Parkinson's Disease dementia is the second most common type of dementia. Over 80% of people living with Parkinson's Disease develop dementia. Identification of an impending dementia is difficult as the clinical presentation of symptoms is unique. Best practice guidelines for cognitive evaluation of people living in PD have been developed to solve this problem. To complement the guidelines, a customised, scalable, and technologically advanced platform, called PDCogniCare, will manage routine cognitive assessments in people living with PD. This project will create an implementation framework for the adoption and translation of PDCogniCare to a national and international scale. The exploration of barriers and facilitators will determine the knowledge transfer required for adoption. Current methods of diagnosis and cognitive evaluation will be examined; recommendations will be made for the roadmap to implementation. |
Expected outcomes and deliverables | Students will have the opportunity to become familiar with literature reviews and the development of clinical guidelines. Contingent on the work completed, students may be eligible for co-authorship on a publication at the conclusion of the program. Students will be expected to produce a brief report of their experience during the program. |
Suitable for | This project is open to students from all disciplines. Students with interest in neurodegenerative disease are highly encouraged to apply. |
Primary Supervisor | Dr Emily McCann |
Instructions to applicants | The supervisor CAN be contacted by students prior to submission of an application. |
Machine learning to detect plasmids from short-read sequences
Hours of engagement | 36 |
Location | Herston: UQ Centre for Clinical Research (918) |
Project description | Plasmids play a key role in gene exchange between bacteria and often carry genes conferring resistance to antibiotics and survival in extreme environments. However, they are difficult to fully characterise from short read whole genome sequencing data alone. This is because plasmids are typically full of repeat sequences that can cause problems for short read assemblers. Long read sequencing can solve this issue, however this technology is not routinely used. Additionally, the majority of sequencing data generated for bacteria to date has been done with short read platforms. We have developed an approach to discover plasmid sharing from short-read sequencing data. This approach relies on a reference database of plasmids, which we organise into a network of "communities" and "sub-communities" using gene jaccard ("Pandora") and rearrangement distance ("Pling"). We use this network to identify shared regions of the graph between multiple short-read sequenced bacteria, in order to infer if two bacteria have potential plasmid sharing. This project aims to apply machine learning techniques to more rapidly predict plasmids within our structured database from short read sequenced bacteria. The machine learning approach will enable us to detect plasmid sharing events more rapidly, reducing the timeframe to appropriate infection control interventions in public health settings. |
Expected outcomes and deliverables | Participants on this project will learn how to navigate UQ's High Performance Computing (HPC) cluster "Bunya", and hone their bioinformatic and command line skills. In addition to learning about bacterial genetics, the participant will learn several key concepts in microbial genomics, including read QC, read mapping, de novo assembly, and alignment. The participant will also gain appreciable knowledge in machine learning techniques and how they can be applied to biological datasets. On completion of the project, the work will be published on a github repository and incorporated into our existing pipeline for discovering plasmid sharing. The results will also contribute to a publication, which the participant will be authored on. |
Suitable for | This project is suitable for a student with experience using command-line and bioinformatic tools who would like to apply their knowledge to a biological problem. |
Primary Supervisor | Dr Leah Roberts |
Instructions to applicants | The supervisor MUST be contacted by students prior to submission of an application. |
Mechanisms influencing impact of environment on mosquito vectors and the transmission of vector-borne diseases across the Pacific
Hours of engagement | 4-5 days per week |
Location | Herston: UQ Centre for Clinical Research (918) |
Project description | This project aims to investigate how environmental factors, including climate variables, land-use changes, and habitat characteristics, influence mosquito vector populations and the transmission of vector-borne diseases (VBDs) across the Pacific. The Aedes polynesiensis mosquito is an important vector in the region, responsible for transmitting several VBDs, such as dengue, Zika, chikungunya, and lymphatic filariasis. Understanding how environmental drivers affect mosquito populations and their ability to transmit disease is crucial for predicting the future risk of VBDs in the Pacific islands, particularly in the context of climate change. Research has shown that factors like temperature, rainfall, and habitat characteristics (e.g., proximity to water bodies, vegetation type, and urbanization) play a significant role in mosquito distribution, survival, and infection rates. Changes in land use, such as urbanisation and agricultural practices, can alter mosquito habitat suitability and increase the potential for disease transmission. With climate change impacting rainfall patterns and temperature extremes, it is important to understand how these factors will influence mosquito populations and the spread of VBDs in the Pacific region. This project will focus on conducting a literature review to identify key environmental drivers of Aedes polynesiensis populations and VBD transmission. The student will also help extract preliminary data to test the relationship between these environmental factors and mosquito abundance, as well as the prevalence of VBDs in the region. Additionally, the project will lay the groundwork for future modelling work to explore the impacts of environmental conditions, such as temperature extremes, on mosquito heat tolerance and vector competence. |
Expected outcomes and deliverables | By the end of the 4-week project, we expect the student to deliver: These outcomes will contribute to a broader understanding of how environmental changes influence the risk of VBDs in the Pacific and inform strategies to mitigate the public health impact of these diseases in the face of climate change. |
Suitable for | We are looking for someone that has a genuine interest in this project and topic. This project is open to applications from students with a background in biology, environmental science, pre-medical provisional students, public health students and students considering a Masters or PhD. |
Primary Supervisor | Drs Helen Mayfield and Eloise Skinner |
Instructions to applicants | The supervisor CAN be contacted by students prior to submission of an application. |
Strategies for enrichment of Mycoplasma genitalium to improve next generation sequencing outcomes
Hours of engagement | 36 |
Location | Herston: UQ Centre for Clinical Research (918) |
Project description | Mycoplasma genitalium is a sexually-transmitted bacterium of global public health significance. It is as common as the leading notifiable STI, chlamydia, however efforts to manage these infections are severely complicated by antimicrobial resistance. Despite the significance of antimicrobial resistance in this public health pathogen, to date there are limited available whole genome sequences of M. genitalium (less than 40 whole genomes globally, with only two of these from Australia). Given this, understanding the molecular mechanisms of resistance to current and emerging treatments is difficult with a lack of genomic data available. Current attempts for whole genome sequencing (WGS) of M. genitalium have been hampered due to the low microbial load of these infections, and the prohibitive expense of RNA baiting approaches to concentrate M. genitalium DNA within clinical patient samples. As a result, only cultured strains of M. genitalium have been subjected to WGS; however, less than optimal culture techniques mean that we are very likely missing/losing the diversity of strains within these clinical samples through the process of culture. This project will focus on optimising a low-cost antibody-mediated enrichment technique for M. genitalium, such that we can concentrate/improve the presence of M. genitalium DNA in clinical samples, for the purpose of improving the success of whole genome sequencing directly from clinical samples. This antibody-mediated enrichment technique has recently been used for another highly fastidious STI, syphilis, and shows great promise for improving the low rates of whole genome sequences for M. genitalium both locally and globally. |
Expected outcomes and deliverables | In this project, students will develop laboratory skills including working with antibodies and mock clinical samples. Students will also assist with data collection, data analysis and contribute towards the publication of this important research. |
Suitable for | This project is suitable for students with a background in biomedical research and/or a strong interest in infectious diseases and molecular diagnostics. Students considering a PhD in infectious diseases are also encouraged to apply. |
Primary Supervisor | Dr Emma Sweeney |
Instructions to applicants | The supervisor MUST be contacted by students prior to submission of an application. |
Understanding the role of intracellular growth during Mycoplasma genitalium propagation
Hours of engagement | 36 |
Location | Herston: UQ Centre for Clinical Research (918) |
Project description | Sexually transmitted infections (STIs) are a global public health concern, responsible for 2.5 million deaths and 374 million new cases annually. Mycoplasma genitalium is one such sexually transmitted pathogen that is as common as the leading notifiable STI, chlamydia, however, efforts to manage M. genitalium infections are hindered by rising rates of antimicrobial resistance and limited treatment options. Cultivation of M. genitalium requires specialised protocols to maintain microbial viability; however, little is understood about the overall growth characteristics and intracellular growth of this public health pathogen. Intracellular growth is a common characteristic whereby pathogens can manipulate host cellular processes and gene expression to promote survival of the microbe. While previous studies have suggested around 10% of M. genitalium strains can internalise within cultured cells; the ability of this microorganism to replicate/survive in host cells, and the factors that drive internalisation, remain poorly understood. This is a major challenge in current laboratory protocols and techniques for cultivation of M. genitalium, resulting in major challenges in successfully propagating and maintaining this microorganism. Optimising culture techniques to address this limitation is crucial for advancing our M. genitalium research, and to increase the number of strains available for future biomedical research (there are currently less than 50 strains available globally). |
Expected outcomes and deliverables | Within this project, students may learn valuable laboratory skills such as cell culture, microbial culture and fluorescent staining/imaging. Students may also assist with data analysis and contribute towards the publication of this important research. |
Suitable for | This project is suitable for undergraduate students with a background in biomedical research with a strong interest in infectious diseases or molecular diagnostics. Students considering a PhD in infectious disease are also encouraged to apply. |
Primary Supervisor | Dr Emily Bryan |
Instructions to applicants | The supervisor MUST be contacted by students prior to submission of an application. |
Visualising a timeline for the elimination of Lymphatic Filariasis in the Pacific Island countries and territories
Hours of engagement | 36 |
Location | Herston: UQ Centre for Clinical Research (918) |
Project description | Lymphatic filariasis (LF) is a debilitating neglected tropical disease that has been targeted by the World Health Organization (WHO) for elimination as a public health problem by 2030. To achieve validation from the WHO that LF has successfully been eliminated as a public health problem, surveillance and monitoring is required to confirm that prevalence levels are below a critical threshold. Countries are officially validated as having eliminated LF as a public health problem after successfully submitted a dossier that includes results from national surveys showing that a country has reduced LF prevalence below critical thresholds needed to interrupt transmission. The Pacific Island Countries and Territories (PICTs) have made significant progress towards LF elimination; 8 of the 21 countries that have achieved validation of the elimination of LF as a public health problem are PICTs. Successful applicants will assist the research team in the development of a dashboard that will summarise pre-validation survey results from PICTs that have achieved elimination. The dashboard will highlight temporal and spatial trends in LF prevalence from the start of elimination efforts to being validated as eliminated. This dashboard will help to identify areas of high LF prevalence that can be used to inform surveillance efforts. Specifically, the successful candidate will be involved in the review data from national survey records, extract data, and help generation a dashboard using PowerBI. Over the course of the project, they will develop a deep understanding of LF and disease elimination in PICTs. |
Expected outcomes and deliverables | Successful applicants will assist the research team in a literature, data extract, and the development of a dashboard that will summarise pre-validation survey results from PICTs that have achieved elimination. The dashboard will highlight temporal and spatial trends in LF prevalence from the start of elimination efforts to being validated as eliminated. This dashboard will help to identify areas of high LF prevalence that can be used to inform surveillance efforts. Specifically, the successful candidate will be involved in the review data from national survey records, extract data, and help generation a dashboard using PowerBI. Over the course of the project, they will develop a deep understanding of LF and disease elimination in PICTs. |
Suitable for | This project is suitable for students with an interest in the data visualisation, spatial mapping, and epidemiology of infectious diseases. The project will involve developing skills in the use of quantitative data, data visualisation, and the use of PowerBI. |
Primary Supervisor | Dr Adam Craig |
Instructions to applicants | The supervisor CAN be contacted by students prior to submission of an application. |