Abstract
RNA metabolism is an important physiological process of organisms. However, almost RNA metabolisms are still unrevealed. The RNA metabolism can be dissected from variation of an expression and an exon structure pattern affected from a condition change. In this research, we present mathematical model to elucidate transcriptome dynamics from multiple RNA-seq data. Expressed mRNA is observed with tag counting after fragmentation in RNA-seq. Then we do not know to which RNA molecular the fragment corresponds. For this matter, correlation information based on multi-conditional RNA-seq data is suitable to model the exon structure and transcriptome dynamics of RNA. Consequently, we elucidate transcriptome dynamics by modeling correlations of multi RNA-seq data at several scales and dissecting the constructed hierarchical model. We adopt proposed model to RNA-seq data of Arabidopsis for 7 conditions. Then we show novel RNA metabolisms and functions of them extracted from the model.
