EUKARYOTIC PROMOTER RECOGNITION BY BINDING-SITES FOR TRANSCRIPTION FACTORS

A method for identification of eukaryotic promoters by localization of binding sites for transcription factors has been suggested. The bindi ng sites for a range of transcription factors have been found to be di stributed unevenly. Based on these distributions, we have constructed a weight matrix of binding site localization. On the basis of the weig ht matrix we have, in turn, designed an algorithm for promoter recogni tion. To increase the accuracy of the method, we have developed a rout ine that breaks any promoter sample into subsamples. The method to be reported on allows much better recognition accuracy than does the appr oach based on detection of the TATA box. In particular, the overpredic tion error is three times lower following our method. The program FunS iteP recognizes promoters from newly uncovered sequences and tentative ly identifies the functional class the promoters must belong to. We ha ve introduced the notion of 'regulatory potential' for the degree to w hich any region of the sequences is similar to the real eukaryotic pro moter. By making use of the potential, we have revealed putative trans cription start sites and extended regions of transcription regulation.