An RNA-Seq Analysis Pipeline for Switchgrass Integrating Co-Expression Network Analysis

Recent developments of high-throughput sequencing biotechnologies enable monitoring thousands of molecules (RNA, protein or metabolites) simultaneously. These technologies have the potential to revolutionize how we study plants and other organisms. For example, RNA-sequencing can show a complete overview of expressed genes (transcriptome) in a cell. The bottleneck, however, is the lack of computational tools and expertise in making effective use of massive datasets generated in the life sciences. Specifically in plants, a serious problem for RNA-seq analysis with plants is mapping uncertainty. With repetitive and/or greater than diploid genomic sequences from outcrossing tissue, one short RNA-seq read sequence can be mapped to different locations with the same weight. In switchgrass (Alamo or many other lowland populations), 25%-40% mapping uncertainty is caused by its tetraploidy nature as well as high level of repetitive sequences of the genome. This study is to develop a much-needed computational pipeline for analyzing RNA-sequencing data in plant genomes. We propose a computational model integrating co-expression module analyses, attempting to reduce mapping uncertainty and therefore increase the power of existing data to identify additional genes that may decrease cell wall recalcitrance.