A New Approach to Sequence Analysis Exemplified by Identification of cis-Elements in Abscisic Acid Inducible Promoters

Identification of cis-regulatory elements in DNA promoters by computational approaches are a great help in genome-wide elucidation of signal transduction, transcriptional regulation and gene expression. However, to accomplish this has proven a difficult problem in computational genomics in part due to the redundant nature of cis-regulatory elements1. Moreover, promoter context and chromatin structure have an important influence on the function of cis-elements and does further complicate the identification of functional cis-elements2,3.
A related computational approach to promoter studies is to use the sequences of known cis-elements for promoter prediction4. This is a computational alternative to genome-wide DNA array analysis of gene expression and can be used to confirm the importance of known promoter elements on a genome level. Unfortunately, this approach does not yield information about new cis-regulatory elements.
We have developed a method for identifying short, conserved motifs in biological sequences such as proteins, DNA and RNA5.
This method was used for analysis of approximately 2000 Arabidopsis thaliana promoters that have been shown by DNA array analysis to be induced by abscisic acid6. These promoters were compared to 28000 promoters that are not induced by abscisic acid.
The analysis identified previously described ABA-inducible promoter elements such as ABRE, CE3 and CRT1 but also new cis-elements were found. Furthermore, the list of DNA elements could be used to predict ABA-inducible promoter with more than 70 % accuracy for the top 40 predictions, which is better than previously described methods4.
In conclusion, identification of short, conserved motifs is a useful computational approach for identification of cis-elements in DNA promoters and for promoter prediction in genome analysis.