Differences

This shows you the differences between two versions of the page.

--- open_projects [2024/01/22 12:24] – project
+++ open_projects [2025/02/28 16:40] (current) – project
@@ Line 4: / Line 4: @@
 If you are a potential supervisor, [[supervisor_instructions:click here]]
+=== Decoding the relationships between DNA replication, genome architecture, chromatin organisation ===
+Contact: Dr Mathew Jones (mathew.jones@uq.edu.au)
+The human genome is packaged into chromatin and assembled into 3D self-interacting chromatin domains that regulate gene expression and coordinate the process of DNA replication. Understanding the relationships between genome structure and function is one of the outstanding challenges in modern biology. Changes in the 3D structure of the genome can cause copying errors (genetic mutations) during DNA replication that results in diseases such as cancer and advanced aging. Decoding the relationships between the genomic landscape and cellular processes such as DNA replication has the potential to inform the development of novel treatments that can treat cancer and extend longevity.
+In this project we are seeking talented and enthusiastic postgraduate students to tackle two fundamental questions: 1. How does the epigenome and the 3D organisation of the genome regulate DNA replication? 2. How are these processes disrupted in cancer and impacted by cancer therapies. The project will assess the impact of genomic features on replication using nanopore sequencing data generated by the Jones lab’s and their artificial intelligence assay for assessing DNA replication in human cells (https://doi.org/10.1101/2022.09.22.509021) and publicly available Hi-C, Repli-Seq, CUT & RUN, ChIP-seq, scSeq, datasets (e.g., GEO, ENCODE).
+Bioinformatics and Computer Science students with skills in R, Python and C++ that are familiar with software suites for the comparison, manipulation and annotation of genomic features are encouraged to contact Dr Mathew Jones (mathew.jones@uq.edu.au) to learn more about the projects available.
+=== De novo identification of insertion sequences with De Bruijn graphs ===
+Contacts: Leah Roberts l.roberts3@uq.edu.au, Tom Stanton tom.stanton@monash.edu
+Insertion sequences (IS) are small DNA elements that can replicate and move throughout a bacterial genome independently. This ability often results in their insertion upstream of or within genes, which consequently leads to large effects on gene expression within the bacteria. These changes in expression can affect a multitude of phenotypes, including resistance to antibiotics and virulence. As such, is it necessary that we characterise where these IS move to within the genome.
+Unfortunately, as IS are small and repetitive throughout the genome, they cause issues for short-read de novo assemblers, most notably collapsed repeats. Because of this, it can be difficult to determine the exact location of IS throughout the genome.
+This project aims to develop an end-to-end pipeline for de novo IS discovery using De Bruijn graphs, and quantify in a collection of bacterial genomes the effect of IS insertions on phenotype.
+This project is suitable for an honours or Masters student. Some background in command line, HPC and python is highly desirable. In this project, you will learn about bacterial genomics and pipeline managers (e.g. Snakemake) in addition to bioinformatic tool development and testing. This project will be based at UQCCR (Herston Campus) and co-supervised by Dr Tom Stanton (Monash University).
+=== Machine Learning to predict plasmids from bacterial isolates ===
+Contacts: Leah Roberts l.roberts3@uq.edu.au, Zamin Iqbal zi245@bath.ac.uk
+Plasmids play a key role in gene exchange between bacteria and often carry gene conferring resistance to antibiotics and survival in hospital 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 which can cause problems for short-reads assemblers. Long-read sequencing can solve this issue, however this technology is currently not routinely used in healthcare settings.
+We have developed a plasmid network that allows users to predict the types of plasmids in their bacterial samples based on gene presence/absence. This project would build upon this work by creating a machine learning framework (Random Forest) that can predict plasmid presence from short-read bacteria.
+This project is suitable for an honours or Masters student. Background in command line, HPC and python is highly desirable. This project will be based at UQCCR (Herston Campus) and co-supervised by Prof Zamin Iqbal (University of Bath, UK).
+=== Pangenomes to predict bacterial transmission in healthcare settings ===
+Contacts: Leah Roberts l.roberts3@uq.edu.au, Michael Hall michael.hall2@unimelb.edu.au
+Predicting whether two bacterial isolates are the same (and thereby inferring if transmission has occurred) has traditionally been performed by identifying and counting single nucleotide variants (SNVs). To do this, a reference genome is usually selected, and isolate reads are mapped to the reference to identify SNVs in regions shared between all isolates. However, for large datasets of very diverse bacterial strains, a single reference genome is usually insufficient, as the shared regions between the strains becomes a very small proportion of the total genomic content.
+We propose a novel method using pangenome reference graphs to better identify and discriminate transmission of bacterial pathogens. This project would start to build test datasets and develop novel workflows for predicting transmission from pangenome graphs.
+This project is suitable for an honours, Masters, or PhD student. Background in command line, HPC and python is highly desirable. This project will be based at UQCCR (Herston Campus) and co-supervised by Dr Michael Hall (University of Melbourne).
+=== Investigation of the effect of the circadian rhythm on the genetic control of gene expression ===
+Contact: Sonia shah <sonia.shah@imb.uq.edu.au>, Solal Chauquet <uqschauq@uq.edu.au >
+The circadian rhythm reflects the daily cycle of behaviours and metabolic processes organisms exhibit. A 24-hour gene expression pattern occurs at the molecular level, with genes activated either during the day or night. Different tissues all display circadian control, with some more affected than others. Within the liver, for example, 3000 genes are subjected to circadian control. This regulation is orchestrated by a small group of CLOCK genes, establishing feedback loops that result in rhythmic gene expression in every tissue.
+We know that gene expression can be influences by genetics variants, called expression quantitative trait loci (eQTL), and this may be one mechanism linking genetic variants to disease. As a result, large eQTL datasets have been generated to assist in understanding disease mechanisms. However, it remains unknown whether sample collection time can affect eQTL identification. This project therefore aims to identify the possible effects of the circadian rhythm on the genetic control of gene expression using the Genotype-Tissue expression (GTEx) dataset.
+During this project, you will run Python tools such as PEER and tensorQTL to identify eQTL within 49 tissues. You will subsequently investigate the associations identified and follow up on the role of the genes under circadian controls within different phenotypes.
+=== Understanding the influence of taste and olfactory perception on eating behaviour and health conditions using big genetic data ===
+Contact info: Daniel Hwang <d.hwang@uq.edu.au>
+Project description: Human perception of taste and smell plays a key role in food preferences and choices. There is a large and growing body of work suggesting that taste and smell (together known as "chemosensory perception") determine eating behaviour and dietary intake, a primary risk factor of chronic conditions such as obesity, cardiometabolic disorders, and cancer. Evidence to date is largely based on observational studies that are susceptible to confounding and reverse causation, leaving the "causal effects" of chemosensory perception on food consumption unclear. If their relationship is truly causal, flavour modification may represent a tangible way of modifying food consumption in a way that benefits public health outcomes. This project aims to: (i) elucidate the genetic architecture underlying individual differences in taste and smell perception, (ii) use this information to assess their causal effects on eating behaviour, and (iii) create a sensory-food causal network mapping individual sensory qualities (i.e. sweet taste, bitter taste, and more) to individual food items.
+=== Increasing drug success rate in human clinical trials using genomics ===
+Around 90% of drug candidates fail in human clinical trials largely due to lack of efficacy or safety concerns. This partly reflects the limitations of using in vitro and animal studies to predict the effect of compounds in humans. Recent studies highlight that drug targets backed by evidence from human genetic studies are 2 times more likely to make it to market. Human genetic data can also identify potential adverse side effects. Such information prior to embarking on human clinical trials could improve the success rate of a compound in human clinical trials and help avoid adverse outcomes for participants. This project will use statistical genomics analyses using publicly available human genomic data to predict efficacy as well as any safety concerns of compounds that are currently in the drug development pipeline.
+Project significance: Findings from this project could potentially identify new therapeutic applications for these compounds or unknown side effects, and ultimately informing future human clinical trials.
+Contact: Sonia Shah <sonia.shah@imb.uq.edu.au>
+Supervisors: You will be working with a multidisciplinary team of supervisors Prof Dave Evans, Dr Sonia Shah, Prof Glenn King, Assoc/Prof Nathan Palpant
+Familiarity with computational analyses (e.g using R or python etc) is needed for this project. Some knowledge around genome-wide association studies and statistical genomics methods such as Mendelian randomisation analysis would be beneficial
+=== Developing quiescent stem cell classifier using single cell transcriptomics ===
+Contact info: Dr Lachlan Harris (Lachlan.Harris@qimrberghofer.edu.au), Dr Olga Kondrashova (Olga.Kondrashova@qimrberghofer.edu.au)
+Quiescence is a reversible state of cell-cycle arrest, sometimes referred to as the “G0” phase of the cell-cycle. It is an adaptive feature of most adult stem cell populations, where it ensures that stem cells divide only when needed, preserving regenerative capacity. However, quiescence is also adopted by cancer stem cells to evade chemo- and radiotherapies that preferentially kill fast-dividing cells. Single-cell data promises to uncover the molecular regulation of quiescent stem cells in health and disease but the identification of these cells within these datasets is either reliant on expert knowledge and manual curation or is currently impossible, due to a lack of marker genes.
+The most common classifiers that define cell-cycle stages (G1/S/G2/M) in single-cell RNA-sequencing data (scRNA- seq) were trained on populations of actively cycling cells. Therefore, these tools cannot identify quiescent stem cells in “G0” phase of the cell-cycle. It is an outstanding question as to whether there are sufficient transcriptomic similarities across quiescent stem cells from different tissue types to build a generalisable model to discriminate these cellular populations. Furthermore, it is unknown whether such a model would generalise to cancerous tissue, where increased variability in transcriptomic states often degrades the distinction between cell types.
+This project aims to develop a broadly applicable quiescent classifier. As a first step towards this, this project will seek to 1) contribute to the curation of datasets and isolation of tissue-agnostic and tissue-specific feature sets that define quiescent stem cells and 2) compare methods for training quiescent classifiers and for determining the most salient features.
+=== Understanding sex-specific cardiovascular disease risk ===
+Contact info: Dr Sonia Shah (sonia.shah@imb.uq.edu.au), Dr Clara Jiang (j.jiang@uq.edu.au)
+Description: Cardiovascular diseases (CVD) account for 35% of female deaths globally (29% in Australia). However, CVDs remain under-studied, under-diagnosed and under-treated in women. This sex disparity is partly due to the lack of knowledge of female-specific risk factors. This project involves statistical analysis of large-scale health and genetic data to identify sex-specific CVD risk factors and underlying mechanisms.
+Requirements: A background in genetics and computational data analysis is preferable.
+=== De-risking the drug development pipeline by finding biomarkers of drug action ===
+Supervisor: Dr Nathan Palpant (n.palpant@uq.edu.au)
+Greater than 90% of drugs fail to advance into clinical approval. Genetic evidence supporting a drug-target-indication can improve the success by greater than 50%. This project aims to make use of consortium-level data resources (UKBiobank, Human Cell Atlas, ENCODE etc) to identify genetic links between genetic targets and phenotypes to help facilitate the translation of drugs from healthy individuals (Phase 1 clinical trial assessing safety) into sick patients (Phase 2 clinical trial assessing efficacy). Finding orthogonal biomarkers of drug action in healthy individuals is critical to de-risk drug dosing when transitioning from Phase 1 to Phase 2 trials. Using ASIC1a as a candidate drug being developed to treat heart attacks, we aim to develop a functionally validated computational pipeline to predict orthogonal biomarkers of ASIC1a inhibitor drug action in healthy individuals to help inform dosing in human clinical trials. Computationally predicted biomarkers will be validated using genetic knockout animals and pharmacological inhibitors of ASIC1a. Collectively, this project will help develop proof-of-principle computational pipeline for orthogonal biomarker prediction of drug targets in the human genome.
+=== Parsing the genome into functional units to understand the genetic basis of cell identity and function ===
+Supervisor: Dr Nathan Palpant (n.palpant@uq.edu.au)
+The billions of bases in the genome are shared among all cell types and tissues in the body. Understanding how regions of the genome control the diverse functions of cells is fundamental to understanding evolution, development, and disease. We recently identified approaches to define diverse biologically constrained regions of the genome that appear to control very specific cellular functions. This project will evaluate how these biologically constrained regions of the genome have influenced evolutionary processes, evaluate their regulatory basis in controlling the identity and function of cells, and analyse the promiscuity of cross-talk between different biologically constrained regions. The project will also study how these genomic regions impact disease mechanisms by evaluating how disease-associated variants in different regions influence survival of patients with cancer and assessing whether these regions are associated with identifying causal disease variants in human complex trait data. The project will involve significant collaborative work with industry partners and researchers across Australia with the goal of providing critical insights into fundamental mechanisms of genome regulation.
+=== Machine learning integration of sequencing and imaging data in cancer research ===
+Cancer is a complex disease that is difficult to treat due to the high level of variation within a tumor and between patients. To better understand cancer complexity at the tissue level, we use a combination of techniques such as single-cell sequencing, spatial transcriptomics, tissue imaging, statistical learning, and deep learning. We analyze the data using high-performance computing to computationally reconstruct biological regulatory networks underlying human diseases in every single cell and between cells within a tissue, like a tumor. By measuring both molecular profiles of the cells and their neighborhood environment, we can integrate genomics and imaging data for earlier and more accurate diagnosis and prognosis of diseases from using tissue biopsies. Our goal is to advance the understanding of biomarkers and cellular regulatory networks that are specific to cell types and tissue microenvironment, which will contribute to early disease diagnosis, targeted drug discovery, and precision medicine.
+Contact: Quan Nguyen <quan.nguyen@imb.uq.edu.au>
 === Resolving molecular trajectories of differentiation pathways of adaptive immune cells in chronic infection generating long-term memory ===
@@ Line 97: / Line 206: @@
 The project aims to:
 .	Compare available machine learning Convolutional Neural Networks (CNNS) algorithms ([[http://kipoi.org|http://kipoi.org]]) to decode the regulatory drivers of cellular ageing.
 .	Statistically associate phenotypic variants from GWAs studies impacting the ageing regulatory lexicon.
@@ Line 106: / Line 216: @@
-=== Trans-ancestry conditional analyses of genome-wide association studies ===
+=== Decoding Transcription Factor Dosage Effects on Cell State Transitions with DoseH-Seq ===
-Contact: Dr Loic Yengo (l.yengo@imb.uq.edu.au)
+Contact info: Dr Christian Nefzger (c.nefzger@imb.uq.edu.au), Ralph Patrick (ralph.patrick@imb.uq.edu.au) and Marina Naval-Sanchez (m.navalsanchez@imb.uq.edu.au)
-The experimental design of genome-wide association studies (GWAS) consists in testing the association between a large number of DNA polymorphisms and a trait of interest. Classically, these associations are tested using a simple linear regression (i.e. one at a time) framework, which cannot distinguish associations from correlated variants. To solve that issue, conditional and joint (COJO) analyses leverage the correlation structure between polymorphisms to identify subsets of variants that are jointly associated with the trait of interest. Current implementations of COJO algorithms can be applied to GWAS performed in individuals of a single ancestry, where the correlation structure between variants is constant; but they cannot yet handle meta-analyses of GWAS from diverse ancestries (e.g. East-Asian, European).
+Cell identity is controlled by different combinations of transcription factors (TFs) that bind to genomic regulatory elements to regulate gene expression. TF activity is not binary in most instances but graded and in response to TF dosage levels (e.g., Naqvi et al., Nat Genet., 2023, PMID: 37024583). For this reason, TFs are strongly enriched for haploinsufficient disease associations (Seidman et al, 2002, J. Clin. Invest. PMID: 11854316; Van de Lee et al., 2020, Trends Genet., PMID: 32451166) and TF dosage and stoichiometry strongly affects reprogramming outcomes (e.g., Polo et al, 2012, Cell, PMID: 32939092; An et al., 2019, Cell Reports, PMID: 31722212). Furthermore, TF dosage effects may also underlie seemingly contradictory effects linked to overactivation of certain TFs in cancer contexts, including of the Nfi family (Becker-Santos, 2017, The Lancet Discovery Science, PMID: 28596133).
-This project aims at developing a COJO algorithm to simultaneously perform variants selection and meta-analyses of multiple GWAS from participants of diverse ancestries. The research will include: (i) developing and comparing algorithms, (ii) testing the impact of violations of model assumptions through simulations and (iii) writing a C++ based software implementing this algorithm. Application of this research can improve our ability to discover genes involved in the susceptibility of common diseases.
+Single-cell RNA+ATAC-seq is a uniquely powerful assay to measure the impact of TF levels on cell regulatory architecture; however, no tools currently exist to directly study TF dosage effects on temporal cell state transitions. To address these gaps, we developed Dosage and Hashtag sequencing (DoseH-seq), an expansion of the 10x Genomics single-nucleus (sn)RNA+ATAC-seq assay that enables sensitive detection of lentiviral perturbations (e.g., TFs) linked to a heterogeneously expressed promoter. In combination with sample hash tagging, multiple temporal, and dosage states, for theoretically any number of genes of interest, can be profiled. This allows detection of TF dosage-dependent effects on temporal cell state transitions, chromatin architecture, co-factor expression, and the rewiring of TF networks at high-resolution. Compatibility with BGI sequencing technology enables the generation of low-cost datasets.We demonstrate the utility of DoseH-seq by tracking the dosage effects of somatic transcription factor, Nfix, during reprogramming towards pluripotency. Contrary to the current dogma, we find that Nfi overexpression can act either as a reprogramming roadblock or as a reprogramming booster, depending on TF dosage and context. These insights may help resolve the TF’s paradoxical role in cancer. DoseH-seq represents a powerful tool for elucidating, and ultimately controlling, both desired and pathological cell state transitions.
-The ideal candidate will have a good understanding of the multiple linear regression model and will be able to efficiently program in R/Python and C++.
+The applicant would help drive method establishment around our novel DoseH-seq technique and support analysis to understand TFs dosage effects with established data sets. Ideal candidate will be able to efficiently program in R or Python. This project is looking for bioinformatics Masters students (ideally 16 units, but we consider 8 unit applicants as well. We also consider PhD students.
+=== Trans-ancestry conditional analyses of genome-wide association studies ===
-=== DNA sequence analysis to investigate why prevalence of adverse effects to ACE inhibitor medication differs across ancestries ===
+Contact: Dr Loic Yengo (l.yengo@imb.uq.edu.au)
-Contact: Dr Sonia Shah (s.shah1@uq.edu.au)
-The angiotensin converting enzyme (ACE) is a component of the renin-angiotensin pathway which regulates blood pressure. It is a target for blood pressure lowering medication (ACE inhibitors). The efficacy and occurrence of adverse side-effects from ACE inhibitor treatment is different amongst difference ancestries.
-This project will analyse exome sequence data of the ACE gene in different ancestries to determine if there are differences in structure across different ancestries, which may explain the ancestry differences in ACE inhibitor adverse effets.
-The ideal candidate will have knowledge and experience in bioinformatics, particularly DNA and protein sequence analysis and analysis of next generation sequence data. Though not necessary, experience with tools such as NCBI BLAST, samtools, vcftools, other sequence analysis packages in R will be advantageous.
+The experimental design of genome-wide association studies (GWAS) consists in testing the association between a large number of DNA polymorphisms and a trait of interest. Classically, these associations are tested using a simple linear regression (i.e. one at a time) framework, which cannot distinguish associations from correlated variants. To solve that issue, conditional and joint (COJO) analyses leverage the correlation structure between polymorphisms to identify subsets of variants that are jointly associated with the trait of interest. Current implementations of COJO algorithms can be applied to GWAS performed in individuals of a single ancestry, where the correlation structure between variants is constant; but they cannot yet handle meta-analyses of GWAS from diverse ancestries (e.g. East-Asian, European).
+This project aims at developing a COJO algorithm to simultaneously perform variants selection and meta-analyses of multiple GWAS from participants of diverse ancestries. The research will include: (i) developing and comparing algorithms, (ii) testing the impact of violations of model assumptions through simulations and (iii) writing a C++ based software implementing this algorithm. Application of this research can improve our ability to discover genes involved in the susceptibility of common diseases.
+The ideal candidate will have a good understanding of the multiple linear regression model and will be able to efficiently program in R/Python and C++.
@@ Line 176: / Line 283: @@
 === Comparing algorithms to estimate polygenic effects ===
-Contact info: Ben Hayes (b.hayes@uq.edu.au), for further information contact Roy Costilla (r.costilla@uq.edu.au)
+Contact info: Ben Hayes (b.hayes@uq.edu.au),
 With the advent of new genomic technologies comes the need to develop new statistical and computational algorithms that can handle large amounts of data in Animal Science. Within the Bayesian paradigm, current methods to estimate polygenic effects for complex traits rely mostly on Gibbs sampling. These approaches are not necessarily scalable to big datasets as the computation time grows more than linearly with sample size. This means that huge computational resources, in terms of RAM memory and/or computing time, need to be used to fit such models. The aim of this project is to compare the performance of alternative Markov chain Monte Carlo (MCMC) algorithms when estimating polygenic effects for complex traits in tropically adapted beef cattle. In addition to Gibbs sampling, at least two MCMC algorithms will be compared: Hamiltonian Monte Carlo and Variational Inference. The student will also learn the basics of Bayesian Statistics and High Performance Computing at UQ.
@@ Line 189: / Line 296: @@
 The project as available on an ongoing basis for honours or Masters of Bioinformatics students, full time.
+=== Develop tools to impute methylation sites from low-coverage sequencing ===
+Contact info: Loan Nguyen (t.nguyen3@uq.edu.au)
+In humans, the methylation state of CpG sites changes with age and can therefore be utilized as an accurate biomarker for aging. In cattle, biological age prediction based on methylation status could provide key information for genetic improvement programs. Additionally, comparing chronological age with biological age (based on methylation status) can provide important information about the stress an animal has been under during its lifetime.
+In this project, students will use cutting edge data sources including reduce representation bisulphite sequencing data, whole genome bisulphite sequencing, long read sequencing and human methylation data to develop a tool to impute methylation sites from low coverage ONT sequence data.
+This project is designed for students who are studying for Masters of Molecular Biology, Masters of Biotechnology, & Masters of Bioinformatics.
+Available for semester 1, 2 and summer
+=== Differential methylated regions related to puberty in Brahman cattle ===
+Puberty is a complex whole-body phenomenon that affects bone growth. In this study, we investigated how puberty in Bos indicus females affects methylation profiles in the epiphyseal growth plate, the cartilage that is essential to bone growth in long bones. Student will analyse nanopore sequencing data of 12 samples (6 pre-puberty and 6 post-puberty) to call methylation and identify the differentially methylated regions between these two groups.
+This project is designed for students who are studying for Masters of Molecular Biology, Masters of Biotechnology, & Masters of Bioinformatics.
+Available for semester 1, 2 and summer
 === CRISPR ===
@@ Line 205: / Line 330: @@
 Available all year, for Master of Bioinformatics students; suitable for one semester, full-time.
-=== The transcriptional landscape of cardiovascular differentiation ===
-Contact info: Nathan Palpant (n.palpant@uq.edu.au)
-Project description: Analysing the transcriptional landscape of cardiovascular differentiation from stem cells at single cell resolution. Stem cells provide a mechanism for generating all cell types of the body. Understanding the mechanisms by which stem cells differentiation into these diverse cell types is central to utilizing them for understanding developmental biology and maximizing their translational potential for cell therapeutics and drug discovery. In this project, you will make use of and develop computational and statistical tools to study the transcriptional landscape of cardiovascular differentiation at single cell resolution. The project will include implementing protocols for quality control analysis, normalization, and clustering, analysing gene networks underlying cell subpopulations, identifying key genetic regulators of cell states, and helping develop novel strategies for studying and analysing single cell RNA-sequencing data to study biological questions.
-Availability, requirements, etc: The project as available on an ongoing basis for honours or masters of Bioinformatics students, full time.
 === Machine learning and data integration in bioinformatics ===
@@ Line 223: / Line 340: @@
 Availability all year, for bioinformatics students with problem-solving skills, Honours or Masters.
 === Reconstruction of ancestral proteins ===