CSIRO develops new cancer classifier

CSIRO statisticians have developed a new molecular classifier tool which is being tested on clinical bone marrow samples from children with acute lymphoblastic leukaemia (ALL) -- the most common cancer in children under 12 -- as a potential diagnostic tool for identifying the disease and its subtypes.

The classifier, known as GeneRave, has narrowed the number of genes required to classify the disease to nine genes, compared to the tens or hundreds of genes used by other classifier systems, said Dr David Mitchell, of CSIRO Mathematics and Information Sciences.

The researchers hope it will become a cost effective diagnostic tool with applications in developing countries where the cure rate for the disease is much lower than in the developed world due to lack of access to simple diagnostic techniques.

The GeneRave classifier is a flexible tool that rapidly sifts through the large amount of data generated by microarray experiments to identify the smallest set of genes that can be used for prediction, Mitchell said. In comparison with many other microarray analysis techniques, CSIRO's system doesn't look at each array point in isolation, but identifies associations between genes.

The Children's Cancer Institute of Australia is now independently evaluating the ALL classifier on a separate set of samples. Mitchell and his team at CSIRO are also looking at applying the technique to other cancers, and other sample types.

"Our objective is to develop a series of cancer diagnostics and prognostics," he said. "GeneRave is very flexible, and can also be used to model a number of responses, such as survival."

According to Mitchell, the technique can be expanded for use in a variety of areas, including diagnostics and prognostics, pharmacogenomics, toxicogenomics, and agricultural applications including linkage trait analysis and breeding.

He said CSIRO was looking to collaborate with different groups on different applications for the technology, through partnerships and joint ventures.