Distribution analysis of 5' splice site-like sequences in human and mouse pre-mRNAs

摘要

The exon definition model accounts for the paring between the 3' splice site and its downstream 5' splice site. It was demonstrated experimentally that 5' and 3' splice-site pairing in metazoan pre-mRNAs occurred in the two distinct steps, In our previous paper, we proposed a subclass method to predict 5' splice sites in mammalian pre-mRNAs, and suggested that several 5' splice site pattern sequences were observed more frequently on the first or second introns than on the other order introns. However, the relation between the 5' splice site sequences and the order of introns is currently under investigation. On the other hand, positional characterization of false positives using computational prediction of human splice sites was presented. In thepresent study, by using a position-tree method to analyze the distribution of the 5' splice site-like sequences in human and mouse pre-mRNAs, we obtained the 5' splice site sequences whose lengths were minimal but were sufficient for specifying the 5' splice sites. Then we investigated the distribution of the 5' splice site-like sequences that were one nucleotide shorter than the obtained 5' splice site sequences. As a result, it was confirmed that the 5' splice site-like sequences had the tendency of the uneven, distribution within pre-mRNAs.