|Ph.D Student||Akerman Martin|
|Subject||Deciphering Splicing Signals from their Genomic Context|
|Department||Department of Biology||Supervisor||Professor Yael Mandel-Gutfreun|
|Full Thesis text|
Alternative splicing is an RNA processing mechanism creating protein diversity in higher eukaryotes. It is regulated by splicing factors that interact with cis-regulatory elements along exons and introns, acting as positive or negative effectors. As a result, genes can express different arrangements of their coding regions.During the first part of my PhD studies I have focused on the regulation of alternative 3’ splice site events, causing variations in exon length as a result of the alternative recognition of the acceptor site. First, we observed a widespread occurrence of tandem acceptors, or NAGNAG motifs, which are alternative acceptor pairs, separated by 3nt and can lead to variations of one amino acid in the protein sequence. We investigated the unique genomic characteristics of alternatively spliced NAGNAG motifs compared to NAGNAG motifs which are constitutively spliced. We found differences between the two groups regarding the nucleotide composition of the NAGNAG motif, evolutionary conservation of the flanking intron and enrichment of cis-regulatory candidates. These unique features suggest that alternative splicing at NAGNAG motifs is a regulated process.Based on these observations we expanded our interest to other alternative acceptors ranging in distances from 3 to 100 nucleotides. To study their regulation we constructed a supported vector machine (SVM) classifier to distinguish alternative 3’ splice sites from constitutive/pseudo splice sites. We trained the SVM with several sequence features and showed that we can successfully discriminate true alternative acceptors from constitutive/pseudo splice sites. In addition we found that the optimal combination of SVM descriptors varies, depending on the distance between the 3’ splice site pairs. Based on these results we hypothesize that the process of splice site selection is influenced by the distance between the competitive splice sites. In the last part of my studies, we developed a new approach for mapping binding sites of known splicing factors which considers both the genomic environment of a single binding site and the evolutionary conservation of the cis-regulatory elements. Using our method we constructed a splicing regulatory network to study regulatory relationships among splicing factors. This network showed a three tier hierarchical structure which significantly correlated with the expression profiles of the splicing factors. This observation strongly suggests that splicing factors coordinately regulate alternative splicing.