Return Home

The aim is to develop software techniques and tools enabling conversion of genetic profiles of tissue into clinically useful knowledge assisting oncologists and pathologists.

What will this research achieve?

Cancer genomics aims to alter the way that people with cancer are diagnosed, staged, and treated. The field is currently in transition: major sites world-wide, including the PMCC, are now generating high-quality datasets involving hundreds of samples with detailed clinical information that have been analysed on complex microarrays (30,000+ elements). This includes arrays that generate different types of information such as gene expression, gene copy number and chemical modification of the DNA (methylation). Such datasets provide an excellent opportunity to develop and apply innovative approaches to data analysis and mining.

What are the key features?

The project involves three streams of activity:

1. Discovery and development of IT methodology for discerning of biologically meaningful and medically useful knowledge from microarray profiles of tissue, with a focus on cancer diagnosis and treatment:

• Learning from a very small number of samples (VSS-learning)
• Kernel machines and regularisation techniques
•  Learning from structured data
•  Feature Selection.

2. Development of practical algorithms and development of software tools for operational testing, demonstration and operational implementation of the discovered solutions.

3. Application of the results to cancer research and clinical oncology. 

 Research team

Staff

Graduate Students

Associated Graduate Students 

Research Highlights