The Spliced EST track shows Expressed Sequence Tags (ESTs) from Genbank that show signs of splicing when aligned against the genome. Requiring splicing reduces the level of contamination in the EST databases drastically at the expense of eliminating many geniune 3' ESTs. Please see also the Human EST track.
Expressed sequence tags are single read (typically approximately 500 base) sequences which usually represent fragments of transcribed genes. Aligning regions (exons usually) are shown as black boxes connected by lines for gap (spliced out introns usually). In full display, arrows on the introns indicate the direction of transcription. ESTs taken from the 3' end of the transcript are drawn in deep red in full mode. 5' ESTs are drawn in deep blue. Beware that the direction arrows on the "introns" are unreliable due to uncertainties in the Genbank annotations, though the sense strand can often be inferred from the GT-AG splice donor/acceptor rule
To make an EST, RNA is isolated from cells and reverse transcribed into cDNA. The cDNA is typically cloned into a plasmid vector, and a read taken from the 5' and/or 3' primer. For most but not all ESTs the reverse transcription is primed by an oligo-dT, which hybridizes with the poly-A tail of mature mRNA. The reverse transcriptase may or may not make it to the 5' end of the mRNA, which may or may not be degraded. In general the 3' ESTs mark the end of transcription reasonably well, but the 5' ESTs may end at any point within the transcript. Some of the newer cap-selected libraries are starting to hit transcription start reasonably well however. Before the cap-selection techniques emerged, some projects used random rather than poly-A priming in an attempt to get sequence distant from the 3' end. These projects were successful at this, but as a side effect also deposited sequences from unprocessed mRNA and perhaps even genomic sequences into the EST databases. (Even outside of the random-primed projects there is a degree of non-mRNA contamination.) Because of this a single unspliced EST should be viewed with considerable skepticism. However since the human 3' UTRs are quite long requiring splicing does eliminate many genuine 3' ESTs.
To generate this track human ESTs from Genbank are aligned against the genome using the BLAT program. When a single EST aligns in multiple places the alignment having the highest base identity is found. Only alignment that have a base identity level within 1% of the best are kept. Alignments must also have at least 93% base identity to be kept.