Discovering regulatory elements in non-coding sequences by analysis of spaced dyads

The application of microarray and related technologies is currently generat ing a systematic catalog of the transcriptional response of any single gene to a multiplicity of experimental conditions. Clustering genes according t o the similarity of their transcriptional response provides a direct hint t o the regulons of the different transcription factors, many of which have s till not been characterized. We have developed a new method for deciphering the mechanism underlying the common transcriptional response of a set of g enes, i.e. discovering cis-acting regulatory elements from a set of unalign ed upstream sequences. This method, called dyad analysis, is based on the o bservation that many regulatory sites consist of a pair of highly conserved trinucleotides, spaced by a nonconserved region of fixed width. The approa ch is to count the number of occurrences of each possible spaced pair of tr inucleotides, and to assess its statistical significance, The method is hig hly efficient in the detection of sites bound by C-6 Zn-2 binuclear cluster proteins, as well as other transcription factors. In addition, we show tha t the dyad and single-word analyses are efficient for the detection of regu latory patterns in gene clusters from DNA chip experiments. In combination, these programs should provide a fast and efficient way to discover new reg ulatory sites for as yet unknown transcription factors.