Thesis Topic Details

Topic ID:
3433
Title:
Visualisation, modelling and analysis of protein phosphorylation from high-throughput mass spectrometry data
Supervisor:
Mike Bain
Research Area:
Bioinformatics, Machine Learning, Visualization
Associated Staff
Assessor:
Sean O'Donoghue
Topic Details
Status:
Active
Type:
Research
Programs:
BINF
Group Suitable:
No
Industrial:
No
Pre-requisites:
--
Description:
Recent developments in biological technology and techniques, such as time-series mass spectrometry, have resulted in new data sets that are too complex to be easily visualised and understood using current "off-the-shelf" methods. The wide-spread adoption of the Web by biologists has resulted in many databases for novel and specific information, such as Phospho Site Plus, a resource for protein phosphorylation and other post-translational modifications which may have an important functional role in diseases such as diabetes and cancer. However, the advent of all this new information is accompanied by a corresponding increase in the time it takes a researcher to sift through such information in order to gain some insight into their data. This project will create a new tool for visualisation of time-series mass spectrometry data on kinase proteins that will integrate multiple information sources on the proteins in the context of known and hypothesized signalling pathways to better enable analysis and prediction of their functional role(s).

First, an appropriate model will be developed for the interaction pathways of these proteins. The representation and dynamics of this model will be informed by, but not restricted to, discrete formalisms including Petri and Boolean networks. The tool should be able to import and export the model in a machine readable format to allow the use of other packages for simulation and analysis, and be able to incorporate meta-data (such as Uniprot IDs) to enable integration of the model elements with other sources of data. Note that modelling may require significant literature research, or in the case of sufficiently complete datasets, machine learning may be employed to refine and complete the pathways, depending on the time available.

Second, although a model of signalling pathways could be created and displayed using existing software, there is no software that is currently suitable for displaying this data in the way the Garvan researchers who produced the data would like to view it. Sean O'Donoghue at the Garvan Institute is creating software (Aquaria) to easily view and investigate proteins and protein-protein interactions. Although Aquaria is primarily a web-based 3D molecule visualisation tool, it also shows protein interactions, locations, time, clusters, graphs, etc. Its strength is combining data from multiple databases in an easy to use and intuitive interface, which allows users to find information relevant to their molecules of interest very quickly. This project will extend Aquaria to deal with the model developed in the first stage of the project. Success of the project will be determined based on feedback from researchers in the James lab and others at the Garvan Institute.
Comments:
--
Past Student Reports
 
No Reports Available. Contact the supervisor for more information.

Check out all available reports in the CSE Thesis Report Library.

NOTE: only current CSE students can login to view and select reports to download.