Knowledge based construction of functional modules for genetic network in Saccharomyces Cerevisiae

Abstract

Methods of module prediction typically involve clustering of group genes and the respective transcription factors. These methods are based on the hypothesis that a gene is regulated by a transcription factor when the expression patterns of the gene and the transcription factor are similar. However, this method is not able to predict the direct target of a transcription factor or its effects on the gene expression. In this study, we propose a new approach that uses data integration in order to predict transcription mechanisms, i. e. the targets of transcription factors and the effects on gene expression. We analyzed yeast ChIPchip data, and DNA microarray data obtained under various physiological conditions. We predicted the functional classification of unknown genes using a Support Vector Machine. We validated our results by comparing with other module prediction programs, and found that our module prediction method shows a higher accuracy than others module prediction programs. © 2011 The Korean BioChip Society and Springer-Verlag Berlin Heidelberg.