Finding regulatory sites from statistical analysis of nucleotide frequencies in the upstream region of eukaryotic genes

We discuss two new approaches to extract relevant biological information on the Transcription Factors (and in particular to identify their binding sequences) from the statistical distribution of oligonucleotides in the upstream region of the genes. Both the methods are based on the notion of a ``regulatory network'' responsible for the various expression patterns of the genes. In particular we concentrate on families of coregulated genes and look for the simultaneous presence in the upstream regions of these genes of the same set of transcription factor binding sites. We discuss two instances which well exemplify the features of the two methods: the coregulation of glycolysis in Drosophila melanogaster and the diauxic shift in Saccharomyces cerevisiae.