Description

This track summarizes alternative splicing shown in the mRNA and EST tracks. The blocks represent exons; lines indicate possible splice junctions. The graphical display is drawn such that no exons overlap, making alternative events easier to view when the track is in full display mode and the resolution is set to approximately gene-level.

To help reduce the noise present in the EST libraries, exons and splice junctions are filtered based on orthologous human transcripts and the frequency with which an exon or intron appears in mouse transcript libraries. Only those exons and splice junctions that have an orthologous exon or splice junction in the human transcriptome or are present three or more times in the mouse transcriptome are kept. Transcripts labeled as mRNA in GenBank are weighted more heavily, reflecting their typically higher quality. This process is similar to that presented in Sugnet, C.W. et al., Transcriptome and genome conservation of alternative splicing events in humans and mice. Pacific Symposium on Biocomputing (PSB) 2004 Online Proceedings.

Methods

The splicing graphs for each genome were generated separately from their native EST and mRNA transcripts using the following process:

• The mRNAs and ESTs were aligned to the genomic sequence using blat. A near-best-in-genome filter was applied such that only alignments with 97% identity over 90% of the transcript and with a score no more than 0.5% lower than the best score were kept.
• The ESTs and mRNAs were oriented by examining the splice sites used in the genomic sequence. Only consensus splice sites, GT->AG and the less common GC->AG, were used to orient the transcripts.
• Alignments were clustered together by sequence overlap in exons. As new splice sites were discovered, they were entered into the graphs as vertices; the exons, introns, and splice junctions connecting them were recorded as edges. Each graph was considered to be a single locus, although they may be fragments of an actual gene structure. The supporting mRNA and EST accessions for each edge were also stored.
• Truncated transcripts were extended by overlap with other transcripts to the next consensus splice site. This prevented the retention of vertices in the graph that were not true splice sites.

After the splicing graphs were constructed independently for both human and mouse, they were mapped to each other using the entire set of genome human net alignments (viewable on the browser as the Human Net track). Only those exons and splice junctions that were common to both or occurred three or more times in the mouse transcript were kept in the splicing graph. When counting the number of times an exon or splice junction was included in the mouse transcripts, those designated as mRNA were weighted more heavily than those designated as EST.

References

For more information on the human net alignments, see Kent, W.J., Baertsch, R., Hinrichs, A., Miller, W., and Haussler, D. Evolution's cauldron: Duplication, deletion, and rearrangement in the mouse and human genomes. Proc Natl Acad Sci USA 100(20), 11484-11489 (2003).

Credits

This annotation was generated by Chuck Sugnet of the UCSC Genome Bioinformatics Group.