DBTSS : DataBase of Human Transcriptional Start Sites and Full-Length cDNA

Although a large number of new genes were predicted from the draft sequence of human genome, the need in cDNA analyses is still unchanged in order to locate correctly the position of genes and to examine their regulatory regions. First, correctly determined transcriptional start sites (TSS) position of a gene enables us to examine its regulatory region more precisely. It is reported that one gene can has multiple TSS, and this phenomena can be explained either by the differential gene expression or rather loose specification of TSSs [1]. In any case, it would be intriguing to find correlations between TSSs and their upstream sequences. Another importance of determining TSSs is that it helps to predict precisely the full-length (or maximum, at least) coding sequence, which is essential to examine the presence of N-terminal sorting signals, for instance. Positional data of transcriptional start sites (TSSs) and full-length cDNA is an indispensable source of biological information. To obtain the full-length cDNAs, we have developed the ‘oligo-capping’ method [2, 3], which is a 5’-end sequencing of full-length cDNAs. On the base of it, novel database, DBTSS [4] presented in this study was constructed. DBTSS will be useful not only as a full-length version of the RefSeq database [5], but also as a resource for analyzing regulatory information in a variety of expression conditions.