Dept. of Medical Informatics and Clinical Epidemiology
Oregon Health & Science University
With the advent of next generation sequencing techniques like RNA-Seq, there is the potential for unbiased transcriptome-wide analysis of gene expression and alternative splicing irrespective of abundance class of the transcript. One of the potential uses of this technology is to help us understand the role that alternative splicing plays in brain region-specific differences. In this study we focused on two RNA-Seq datasets derived from the mouse brain, one from the striatum and the other from the whole brain. We first quantified the abundance of the different forms of alternative splicing events using a very conservative approach utilizing exon definition information from both the Ensembl and ASTD public databases. We then applied a measure that quantified transcript isoforms and examined whether biases existed in these quantities when stratified by overall gene expression measured using a microarray. Further, we explored whether there was concordance between alternative splicing events quantified using RNA-Seq and those measured using a statistical model for an Affymetrix exon array experiment. Finally we examined whether a simple model-based strategy could be pursued in order to detect alternative splicing in a high confidence dataset and explored some of the properties of this model using simulation. Overall, we found that a major confounder for many of our analyses was the lack of sample size. This is an issue that will be explored further in future work.
School of Medicine
Bottomly, Daniel W., "An in silico assessment of alternatively spliced isoforms in the mouse brain using RNA-Seq" (2009). Scholar Archive. 380.