Computational approaches to delineate non-canonical splicing events

Abstract: RNA splicing generates enormous variations at the RNA and protein levels to regulate cell-type- specific functions as well as being the cause for numerous diseases. Several classes of splicing patterns have been widely studied, such as exon skipping, intron retention and alternative splicing sites. Advances in sequencing technologies enable the discovery of previously unknown non-canonical splicing events. However, due to the lack of dedicated computational approaches, the prevalence and functional consequences of these non-canonical splicing events remain unexplored. The goal of our research program in the next five years is to develop novel and specialized computational algorithms for discovery and characterization of emerging splicing patterns that are currently understudied. We will focus on exitrons and non-linear- spliced transcripts in our proposed study, as these two non-canonical splicing models have been implicated in complex human diseases reported by recent studies. We will develop a series of algorithms to (1) comprehensively catalog these novel splicing patterns using short and long read sequencing platforms, (2) dissect the genetic basis of non-canonical splicing events with integrative analysis of deep transcriptome and whole-genome sequencing data, and (3) elucidate the functional impacts of novel forms of RNA splicing alternations using a proteogenomic strategy. Our proposed work is innovative in that we will build unique computational frameworks to detect and characterize novel non-canonical splicing events by integrating large multi-omics datasets (e.g. TCGA, GTEx and ENCODE). It is significant because it can be applied both in basic research to improve transcriptome annotation and potentially in genomic medicine to guide the development of novel therapeutic strategies for complex diseases.