Dissecting Eukaryotic Transcriptomes Through High-Throughput Data

• Activity Number: 527
• Type: Topic Contributed
• Date/Time: Wednesday, August 7, 2013 : 10:30 AM to 12:20 PM
• Sponsor: Biometrics Section
• Author(s): Liang Chen*+
• Companies: Molecular and Computational Biology, University of Southern California
• Keywords: RNA-seq ; alternative splicing ; isoform quantification ; , transcriptome comparison ; splicing regulatory elements

Abstract

High throughput technologies have enabled us to dissect transcriptomes at an unprecedented resolution. However, efficient and robust statistical methods are still in urgent need. We proposed a two-parameter generalized Poisson model to address the positional bias in RNA-seq and separate real expression signals from sequencing bias. It also significantly improves downstream analyses such as normalization across different samples, and the identification of differentially expressed genes and differentially spliced exons. To precisely deduce gene regulation, two transcriptomes need to be quantified and compared at the transcript isoform level in additional to the gene expression level, because alternative splicing dramatically increases the complexity of eukaryotic transcriptomes. We developed novel statistical models to quantify isoform expression or directly infer the differential expression level of each transcript isoform. We have also developed a statistical approach to identify functional splicing regulatory elements based on RNA-seq data and incorporated other non-motif based information. All these facilitate the reconstruction of alternative splicing regulatory networks.