A Comprehensive Method and Tool for Identifying Conserved Cis-Element Motifs on the Basis of Large-Scale Gene Expression and Sequence Data

The elucidation of gene expression regulatory mechanism facilitates improvement in agronomical traits of crops. To comprehensively and efficiently identify cis-elements in the genome, we have developed a novel mining algorithm and bioinformatics tool so far. The same cis-elements are frequently shared across genes with similar expression profiles because the same regulatory mechanism may be involved in the controlling their transcription activation. With a list of gene names (e.g. genes with similar expression profiles), our method provides strong candidates of cis-elements which are significantly overrepresented among the up/downstream regions of gene. The current version of our tool can be executed against the genomes of Arabidopsis, rice, soybean, tomato, potato, poplar, grape, Medicago truncatula.

Our tool has three steps to identify candidates of cis-elements. In the first step, frequencies of each nucleotide sequence patterns (7~11bp) in the up/downstream regions (each 3 kb in the default setting) of genes in the genome are acquired. The frequencies are employed as the reference (control). In second step, similarly, the frequencies of each nucleotide sequence patterns are also calculated in the up/downstream regions of target genes (e.g. genes with similar expression profiles). In the final step, the frequencies are compared with our statistical method. Verification of the computational predictions against genes with similar expression profiles in Arabidopsis thaliana (Tatematsu et al. 2005 Plant Physiology) showed high predictive accuracy. The tool will be available from our web-site (http://bioinf.mind.meiji.ac.jp/lab/).

This work is supported by Research Funding for Computational Software Supporting Program in Meiji University.