DBTSS: database of transcription start sites, progress report 2008

Abstract

DBTSS is a database of transcriptional start sites, based on our unique collection of precise, experimentally determined 5'-end sequences of full-length cDNAs. Since its first release in 2002, several major updates have been made. In this update, we expanded the human transcriptional start site dataset by 19 million uniquely mapped, and RefSeq-associated, 5'-end sequences, which were generated by a newly introduced Solexa sequencer. Moreover, in order to provide means for interpreting those massive TSS data, we implemented two new analytical tools: one for connecting expression information with predicted transcription factor binding sites; the other for examining evolutionary conservation or species-specificity of promoters and transcripts, which can be browsed by our own comparative genome viewer. With the expanded dataset and the enhanced functionalities, DBTSS provides a unique platform that enables in-depth transcriptome analyses. DBTSS is accessible at http://dbtss.hgc.jp/.

Figure 1.

Screenshot from the Solexa sequence viewer. Its basic utility is similar to that of the previous version (5). Users can choose the database to search, either Sanger or Solexa dataset, in the left panel and then retrieve the results (red circle). Users can also switch the browsers between the Sanger and Solexa results (blue circle).

Figure 2.

Screenshot from the search engine for enrichment of the putative transcription factor binding sites (A). The figure exemplifies the search for common TF binding sites appearing in the promoters of genes with which more than 10 Solexa sequences are associated and the relative expression levels are more than 2-fold elevated both by the 1 μm trichostatin A treatment and by the 1 μm wortmannin treatment (B). From the resultant list of the enriched sites, the link can be followed to the main viewer to retrieve further detailed information (C).

Figure 3.

Screenshot from the search engine for the evolutionary conservation of the promoters and transcripts (A). The figure exemplifies the results of the search for promoters for which more than 10 Solexa sequences are associated, the alignable region in the promoters in human–mouse comparison is <300 bp and the overall base substitution of the downstream transcript region is <20% (B). The regions specified between red (one selection) and green (second selection) vertical lines (or one left click) can be magnified up to the sequence level.