Financing: Research Foundation - Flanders (FWO)

Project reference Nr.: G.0115.01
Start: 2001-01-01
End: 2004-12-31

Description:
For cancer, the process of diagnosis, staging, prognosis estimation, ... is usually based on empirical criteria. However, the phenotype of the tumor is determined by the collection of disturbed expression levels. By including these expression levels in the clinical decision-making we can take into account the fundamental mechanisms unerlying carcinogenesis. Microarrays allow us to measure these expression levels. However, each microarray experiment produces thousands of values. To gain clinically relevant results we have to use algorithms designed to extract data useful for the medical doctor (data-mining). There are three major tasks to be accomplished by these algo- rithms: 1. Feature selection (detection of relevant genes). 2. Classification of patients. 3. Class discovery (clustering).
 

SMC people involved in the project:

• Bart De Moor (Promoter)