Skip to Main Content
A significant roadblock to the use of genomic data for understanding gene networks in infectious pathogens is our inability to assign functionality to a large fraction of the genes. Nowhere is this more problematic than in the malaria parasite Plasmodium falciparum, in which 60% of the genes are annotated as "hypothetical". To circumvent this problem we proposed to employ wavelets, feature extraction, kernel based supervised learning, and pattern recognition algorithms to explore temporal expression profiles from the complex and dynamic developmental cycle in the parasite and discover crucial network components.
Date of Conference: 28-30 May 2006