DNA regulatory motif selection based on support vector machine (SVM) and its application in microarray experiment of Kashin-Beck disease

Abstract

Conserved DNA sequences are essential to investigate the regulation and expression of nearby genes. The conserved regions can interact with certain proteins and can potentially determine the transcription speed and amount of the corresponding mRNA in gene replication process. In this paper, motifs of coexpressed genes of microarray experiments were explored with discovery algorithms. Then a selection algorithm based on support vector machine (SVM) was applied to identify those motifs which mostly influenced gene expression. This method combined the advantages from both matrix based motif finding and functional motif selection. When applied to Kashin-Beck disease (KBD), this method identified 9 motifs, and revealed that some motifs may be related to the immune reactions. In addition, we suggested that the methods used could be applied to other microarray experiments to explore the underlying relationships between motif types and gene functions.

Key words: Support vector machine (SVM), microarray, motif discovery, gene regulation, Kashin-Beck disease.