Below are projects for 2016-2017.

Investigating Complex Patterns in Lymphocyte Receptor Usage in Immune Mediated Diseases

Project duration:

6-10 weeks depending on applicant

Description:

The cells of the immune system use an array of chemical signals and physical interactions to engage with their environment, enabling them to monitor the health of the host organism and detect instances of infection. In immune mediated diseases (IMDs), inappropriate immune activation can cause these cells to react against components of an individual’s own body, causing significant damage in locations as varied as the brain (multiple sclerosis), joints (rheumatoid arthritis) and gastrointestinal tract (Crohn’s disease). One of these cell subsets, the natural killer (NK) cells, are responsible for targeting and killing virally invaded cells in the context of infection, but have been implicated in many IMDs due to misdirected cytotoxic activity. The activity of NK cells is regulated through surface expression of a range of killer immunoglobulin-like receptors (KIRs) that are encoded in haplotypes varying from 4-20 genes in human populations. KIRs recognise human leukocyte antigen (HLA) molecules that are expressed on every nucleated cell in the body, with different KIR receptors specific for different HLA alleles. The combination of KIR and HLA alleles carried by an individual has been shown to alter propensity for certain IMDs or infections diseases. 

We have available a range of large genotyped disease cohorts for which KIR haplotypes and HLA alleles can be imputed. Statistical analysis to investigate patterns in KIR and HLA inheritance will be highly informative in regards to the role of NK cells in IMDs.

Expected outcomes and deliverables:

1) Understand how to process genetic data (SNP imputation) 2) Learn statistical methods to interrogate patterns in biological data 3) Learn about important concepts in immunogenetics 4) Contribute to journal articles summarising outcomes.

Skillsets that you will improve: R programming, linux and softwares to handle large-scale genetic data sets, statistical analysis of genetic data, visualisation, critical thinking. 

Suitable for:

UQ Masters students in Bioinformatics or Biostatistics, preferably 8-unit course.

Primary Supervisor:

 Dr Kim-Anh Lê Cao, UQDI, Translational Research Institute, in collaboration with Prof Matt Brown and Ms Aimee Hanson (PhD student), QUT

Further info:

k.lecao@uq.edu.au
aimee.hanson@uq.edu.au

Improving the mixOmics R toolkit

Project duration:

6-10 weeks depending on applicant

Description:

The mixOmics R package (http://www.mixOmics.org/) which is an open-source tool we created in 2009 with strong uptake in the research community (> 10,000 R CRAN unique IP downloads in 2015 compared to 4000 in 2014). The mixOmics R library currently includes 15 multivariate methodologies, for single ‘omic analysis, integration of two data sets and more. Some of the methods (sPLS, sPLS-DA) have been applied to address many diverse research questions including recent exciting applications to global-scale ocean genomics (Guidi et al., Nature 532, 465–470 28 April 2016) and to gut microbiota in bacterial infections (Ramanan et al, Science 352, 608-612 29 Apr 2016) published in high impact factor journals. We believe that mixOmics will open new avenues in integrating multiple high and low throughput experiments performed on the same biological samples and has strong potential to identify highly relevant biomarker candidates.
 

We propose the two following projects related to mixOmics:

Project 1. We will further investigate some methodological aspects in the current methods, such as including meta-data in the multivariate models, or choosing the optimal number of components in our multivariate methods. We will apply our novel improvements to data generated in-house, ranging from `omics: proteomics, transcriptomics, 16S microbiome, and package the methods in mixOmics if appropriate. This project is in collaboration with UQDI researchers who generated the data.

Project 2. We will further test and apply a novel method currently developed by the mixOmics team to analyse genotyping data (Single-nucleotide polymorphism, SNP) and integrate them with other types of `omics data (e.g. lipidomics, transcriptomics). This project is in collaboration with a team from UCLA, Los Angeles studying Atherosclerosis in a large cohort of patients.
 
Project 3. We will focus on improving computations and visualisation in mixOmics, investigating Apache Spark and MapReduce for efficient parallelisation, as well as plotly for web-based interactive graphs. The project will greatly improve the package for other worldwide scientists to use. Demonstrated level of R programming is essential for this project. 

Expected outcomes and deliverables:

1) Learn about statistical data integration & multivariate methods implemented in mixOmics 2) Implement new R code functions 3) Test methodologies on existing `omics and genotyping data 4) contribute to the mixOmics project.

Skillsets that you will improve: R programming, statistical analysis of large biological data including application of multivariate methods, statistical learning, data visualisation, critical thinking.
 

Suitable for:

UQ Masters students in Bioinformatics or Biostatistics, preferably 8-unit course.

Primary Supervisor:

Dr Kim-Anh Lê Cao (mixOmics core team), UQDI, Translational Research Institute;
 

Further info:

k.lecao@uq.edu.au

Determining the role of the tumour associated fibroblasts in skin cancer development

Project duration:

6 weeks

Description:

This project will determine the role of the associated fibroblasts in the development of the epithelial tumour, Basal Cell Carcinoma (BCC) of the skin. We aim to identify novel regulated pathways to target for development of new BCC therapeutics. We have identified a novel list of targets, and the project entails doing basic protein or RNA expression profiling to validate our putative targets.  

Expected outcomes and deliverables:

The student will learn basic techniques in tumour characterisation, basic cancer genetics and depending on skill and prior training will observe or perform some microscopy and molecular biology.

Suitable for:

Students with an interest in cancer biology and molecular genetics, with an interest to doing higher research in the future.
 

Primary Supervisor:

Dr. Rehan Villani

Further info:

Email: r.villani@uq.edu.au
Phone: +61 (0) 7 3443 7074
  

Immunotherapy for type 1 diabetes

Primary Supervisor:

Dr Emma Hamilton-Williams

Further info:

 e.hamiltonwilliams@uq.edu.au

Development of oral drug delivery strategies for tolerizing dendritic cells for rheumatoid arthritis therapy

Project duration:

6-10 weeks

Primary Supervisor:

Dr. Ranjeny Thoms, Dr Meghna Talekar

Further info:

m.talekar@uq.edu.au

Assessing trafficking of liposomal systems using multi-photon imagin in rheumatoid arthritis

Project duration:

6-10 weeks

Primary Supervisor:

Dr. Ranjeny Thoms, Dr Meghna Talekar

Further info:

m.talekar@uq.edu.au

Quantitative Analysis of Epigenetic Mutations for Cancer Diagnosis

Project duration:

10 weeks

Description:

 Cancer is characterized by the accumulation of somatic variants at both the genetic and epigenetic levels in the tumour genome. Some evidence that suggests that epigenetic mutations - changes in DNA methylation, specifically - may bear more diagnostic potential for cancer than genetic mutations. We are developing an assay that can detect epigenetic mutations in tumour DNA with extremely high sensitivity and specificity. We want to explore the potential of this assay as a predictor of disease characteristics such as cancer subtype, disease aggressiveness, and response to therapy in the blood cancers lymphoma and myeloma. This project will involve analysis of targeted sequencing data from bisulfite-converted tumour DNA, comparison between normal and cancer sequencing data, and inference of epigenetic signatures associated with disease characteristics.

Expected outcomes and deliverables:

- a catalog of genomic regions that have a strong differential methylation signal in cancer cells relative to normal cells across a range of samples
- statistical inference of DNA methylation signatures associated with cancer subtype or other relevant clinical attributes
- as time and experience allow, development of an R package/Python script/etc. to facilitate future analyses
 

Suitable for:

 Students with programming experience and a background in quantitative science (maths, statistics, ITEE, bioinformatics, etc.) and an interest in applying their knowledge to biology or medicine

Primary Supervisor:

Dr Lynn Fink
 

Further info:

l.fink@uq.edu.au

Deciphering gene networks of mental health disorders

Project duration:

 6-10 weeks

Description:

Mental health disorders such as Autism Spectrum Disorders (ASD), Attention Deficit Hyperactivity Disorder (ADHD), and Schizophrenia (SZ) are highly heritable. They are also highly polygenic complex disorders sharing some genetic risk factors and functional pathways which represent a challenge for an accurate and efficient diagnosis. Despite thousands of genome sequences have been made available through international consortia and public databases very little is known of the cellular and molecular mechanisms through which these genetic factors associate with mental disorders. In previous work we have developed a computational model that integrated genetic data and molecular networks to predict gene set associations with gene networks for four mental disorders (Cristino et al. 2014 Mol Psychiatry). Although we found informative mutations in the protein-coding regions, the great majority (~90%) of these mutations lies in non-coding regulatory regions. However it is still unclear how mutations perturb molecular pathways and cellular processes underlying these complex diseases. We aim to use bioinformatics and systems biology approaches to reveal the cellular and molecular mechanisms disrupted by genetic mutations within the regulatory regions of the genome.  

Expected outcomes and deliverables:

 1) Learn bioinformatics analysis of whole-exome and whole-genome sequencing data; 2) Learn DNA motif analysis to discover transcription factor and microRNA binding sites; 3) Learn basic concept of gene network analysis; 4) Contribute to scientific publication.

Skillsets that you will improve: Python programming, R programming, gene networks inference and analysis, data visualisation, critical thinking.
 

Suitable for:

 UQ Masters students in Bioinformatics or Biostatistics, preferably 8-unit course.

Primary Supervisor:

Dr Alex Cristino
 

Further info:

a.cristino@uq.edu.au