Research on the Characterization and Differential Analysis of the mRNA Transcriptome

"Deep sequencing" technologies are a valuable new tool for biomedical research and medical diagnostics, and provide new opportunities to characterize the transcriptome with unprecedented resolution (Mortazavi, et al. 2008, Wang et al. 2009). Rapid improvements in sequencing quality and sampling depth coupled with efficient new algorithmic techniques enable mRNA splicing structures and expression levels to be acquired automatically and at reasonable cost. How to efficiently analyze and make biological sense of these data is the new challenge.

Our research in this area focuses on the analysis of the mRNA transcriptome, including the separation of RNA sequencing data into its constituent transcripts, the differential analyses of mRNA transcriptomes in experimental samples (e.g. transformed tumor cells vs. healthy cells), and the ability to link transcriptional data to genomic DNA and the predicted proteome. Our approach captures the mRNA transcriptome in the form of an expression-weighted splice graph (ESG) that represents the transcribed portions of the genome, their splicing structure, and the steady state level at which these transcribed regions are expressed. This compact representation is suitable for storage and indexing and is an efficient basis for all computational methods of analysis.

The overarching goal of our approach is to create a data-driven construction and analysis of the ESG, removing dependence on genomic libraries and gene structure annotations. This enables a highly efficient and sensitive detection of previously unknown splicing structures. It also allows the application of RNA sequencing methods to species beyond model organisms, providing the opportunity to investigate novel functions in any species that may have important implications for human health. Our construction and analysis pipeline extends to compltely novel settings where the genome structure might profoundly deviate from a reference genome (such as in cancer).



		Research on the Characterization and Differential Analysis of the mRNA Transcriptome "Deep sequencing" technologies are a valuable new tool for biomedical research and medical diagnostics, and provide new opportunities to characterize the transcriptome with unprecedented resolution (Mortazavi, et al. 2008, Wang et al. 2009). Rapid improvements in sequencing quality and sampling depth coupled with efficient new algorithmic techniques enable mRNA splicing structures and expression levels to be acquired automatically and at reasonable cost. How to efficiently analyze and make biological sense of these data is the new challenge. Our research in this area focuses on the analysis of the mRNA transcriptome, including the separation of RNA sequencing data into its constituent transcripts, the differential analyses of mRNA transcriptomes in experimental samples (e.g. transformed tumor cells vs. healthy cells), and the ability to link transcriptional data to genomic DNA and the predicted proteome. Our approach captures the mRNA transcriptome in the form of an expression-weighted splice graph (ESG) that represents the transcribed portions of the genome, their splicing structure, and the steady state level at which these transcribed regions are expressed. This compact representation is suitable for storage and indexing and is an efficient basis for all computational methods of analysis. The overarching goal of our approach is to create a data-driven construction and analysis of the ESG, removing dependence on genomic libraries and gene structure annotations. This enables a highly efficient and sensitive detection of previously unknown splicing structures. It also allows the application of RNA sequencing methods to species beyond model organisms, providing the opportunity to investigate novel functions in any species that may have important implications for human health. Our construction and analysis pipeline extends to compltely novel settings where the genome structure might profoundly deviate from a reference genome (such as in cancer).
Copyright 2011 Last Modified 4/2011