Differences

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

--- open_projects [2024/10/17 16:18] – project
+++ open_projects [2025/10/17 17:30] (current) – project
@@ Line 5: / Line 5: @@
 If you are a potential supervisor, [[supervisor_instructions:click here]]
-=== De novo identification of insertion sequences with De Bruijn graphs ===
+=== The Barrier Atlas: Cross-Tissue Insights into Homeostasis and Dysfunction ===
-Leah Roberts: l.roberts3@uq.edu.au, Tom Stanton: tom.stanton@monash.edu
+Contact: Amanda Oliver (Amanda.Oliver@qimrb.edu.au)
-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.
+Single-cell RNA sequencing (scRNA-seq) has transformed our understanding of human tissue biology, revealing cellular diversity across organs and disease states. Building on existing datasets profiling millions of cells, this project aims to construct a unified single-cell atlas of barrier tissues, including the lung and gut, to uncover shared and tissue-specific mechanisms that maintain immune balance at the body’s environmental interfaces. The student will develop and apply computational pipelines for large-scale data integration, quality control, cell type annotation, and spatial and microbial mapping across millions of cells and thousands of samples. Advanced methods such as gene regulatory network inference, deep learning, and foundation models will be used to explore cross-tissue immune regulation and barrier dysfunction. By combining single-cell, spatial, and microbiome data, the project will deliver the first cross-tissue atlas of barrier biology, providing new insights into diseases such as inflammatory bowel disease and chronic respiratory disorders.
-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.
+Suitable for Masters, or PhD students. Strong bioinformatics skills using Python or R are essential; experience with single-cell or spatial transcriptomics and knowledge of immunology or barrier tissue biology is highly desirable.
-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.
+=== The Escape of Human Genomic Data into Public Repositories ===
-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).
+Contact: Michael Hall (michael.hall1@uq.edu.au)
+Public sequencing repositories (e.g. SRA) are growing rapidly, but many studies involving human clinical samples may inadvertently include identifiable host DNA—even when ethics approvals explicitly prohibit this. This project investigates the extent and implications of such data leakage.
-=== Machine Learning to predict plasmids from bacterial isolates ===
+Objectives:
+	•	Identify publicly available datasets from clinical pathogen/metagenomic sequencing studies
+	•	Quantify residual human genomic content using a variety of approaches and references
+	•	Benchmark human read detection approaches (e.g. host depletion vs k-mer-based methods)
+	•	Assess potential identifiability using forensic markers (e.g. Illumina Infinium SNPs, CODIS loci)
+	•	Explore the role of ethics language, technical variability, and population bias (e.g. African vs European genomes) in leakage rates
-Leah Roberts: l.roberts3@uq.edu.au, Zamin Iqbal: zi245@bath.ac.uk
+Skills you’ll gain:
+	•	Handling and processing large sequencing datasets
+	•	Working knowledge of alignment and k-mer classification tools (e.g. minimap2, kraken) and human read detection pipelines
+	•	Experience in reproducible bioinformatics analysis and privacy-aware genomic research
+	•	Insight into the intersection of ethics, bioinformatics, and public data governance
-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.
+This project is ideal for students interested in clinical genomics, privacy, ethics, or data-driven policy impact. Familiarity with the command line is necessary. Knowledge of Python would be great, but not required—we can build those skills as you go!
-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.
+=== Decoding the relationships between DNA replication, genome architecture, chromatin organisation ===
-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).
+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.
 === Pangenomes to predict bacterial transmission in healthcare settings ===
-Leah Roberts: l.roberts3@uq.edu.au, Michael Hall: michael.hall2@unimelb.edu.au
+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.