In recent months, several groups published papers about predictive identification of alternative exons from genomic sequences. These papers come from Rotem Sorek, Brenton Graveley and Chris Burge lab.
These work utiltize features that are characteristic of functional alternative exons. For example, alternatively spliced exons are in general shorter. They have slightly weaker splice sites, and are more likely to be in frame. The most important discriminative features probably come from comparative genomics. For example, functional alternative exons are usually spanned by highly conserved intronic sequences.
Posted by yxing at February 12, 2005 01:12 PMThese works didn't use ESTs to detect AS, instead they use comparative genomics and intrinsic signals lying in the genmoes. However, spliceosome does not check the genome of another organism when it is working. Thus, the information in the single genome should be adequate for the purpose of predicting AS. But until now there seems to be no published paper on this topic.
Posted by: Thomas Bee at February 20, 2005 09:01 PM