Abstract
A. Prof. Christian Schoenbach, Nanyang Technology University (NTU), Singapore.

Alternative splicing is an important mechanism that contributes to the phenotype diversity of mammalian cells. Up to 60% of mouse multi-exon genes are alternatively spliced. Yet, the regulation of splicing, that is when and where an exon is skipped or an alternative splice site is used, remains to be fully decoded. Splice signals flanking alternative exons are often weaker because they deviate from the consensus splice motifs. However exonic splice enhancers (ESE) can offset the weaker splicing signals. ESE are 6 to 12-mer motifs that are recognized by spliceosomal proteins. Position weight matrix-based probabilistic modeling is one of many techniques that has helped to identify ESE in alternative exons. We were interested in identifying without prior assumption on splice site strength and in a position-independent way motifs that may differentiate alternative exons from skipped or constitutive exons. Considering the similarities to detecting cis-regulatory elements such as potential transcription factor sites we applied a word enumeration method and an expectation maximization method to a data set described by Wang et al. (2004). Using the word enumeration method we identified 400 mouse (400 human) 6-mer motif candidates. Of these 37% (48%) were unique new exon-type specific candidates. Forty-five percent of all mouse (42% human) 6-mers corresponded to known ESE including experimentally confirmed motifs. Results obtained with the expectation maximization method produced between 46% to 71% (44%-68%) mouse (human) novel splice regulatory motif candidates. The functionality and strength of the motifs and the relationships between motif strength and motif occurrence in particular exon types remains to be examined more closely in future experiments. However, recent findings reported in the literature supported the potential role of some of our candidates in splicing regulation. For example, mouse alternative exon-specific motif CGTGAG is identical to ASSET-reported splicing tag associated with intron retention.
Host: Dr. Frederic Bard.
|