Computational methods for the analysis of RNA-Seq data

University Of Wisconsin-Madison, Madison WI

Investigators

Linked publications & trials

Abstract

DESCRIPTION (provided by applicant): Technologies for the measurement of mRNA quantities within cells are key components of a biomedical researcher<s toolbox. The characterization of gene expression is important to both the understanding of the molecular biology of the cell and the diagnosis and treatment of human disease. To be most useful to scientists, RNA measurement technologies should be as accurate and precise as possible since even small perturbations in transcript levels may be significant. A recently developed experimental method, RNA-Seq, is promising to revolutionize gene expression analysis and is enabling new discoveries about the human transcriptome. RNA-Seq data demands a significant amount of computation before it can be used and the current computational methods are still in their infancy. We propose to take RNA-Seq computational methods to the next level, increasing both the accuracy of gene expression estimates and the number of scenarios in which it may be used. Using novel probabilistic models and statistical learning techniques, we will enable the technology to precisely measure alternative splicing events and characterize the transcriptomes of non-model organisms. Our computational methods will be validated with both real and simulated RNA-Seq data and will be made freely available as an open source software package. In addition, we will use the methods we develop to explore differences between the transcriptomes of undifferentiated and differentiated cells. A first application will be the characterization of alternative splicing differences between human embryonic stem cells and differentiated fibroblast cells. A second application will be the estimation of gene expression levels in embryos of the frog Xenopus leaves using the genome sequence of a closely related frog;Xenopus (Silurana) tropical is as a reference. The results of these experiments are expected to advance our understanding of cellular differentiation in vertebrates and, ultimately, the potential for stem cells to be used in the treatment of human diseases and injuries. PUBLIC HEALTH RELEVANCE: The proposed research aims to develop computational methods for the support of a technology that measures the quantities of RNA inside of a cell. With this technology and the developed computational methods, researchers will be able to better diagnose and understand the molecular basis of human disease.