A Nile Grass Rat Transcriptomic Landscape Across 22 Organs By Ultra-deep Sequencing and Comparative RNA-seq pipeline (CRSP)

Abstract RNA sequencing (RNA-seq) has been a widely used high-throughput method to characterize transcriptomic dynamics spatiotemporally. However, typical RNA-seq data analysis pipelines depend on either a sequenced genome or corresponding reference transcripts or them both. This restriction makes the use of RNA-seq for species lacking both of sequenced genomes and corresponding reference transcripts challenging. Nile grass rat ( Arvicanthis niloticus ) is a diurnal rodent species with several unique characteristics making it as a useful model to study diet-induced type 2 diabetes and other physiological or behavioral processes due to its diurnal nature. However, there is neither a high-quality annotated Nile grass rat genome nor a reference transcript sets available so far, making it technically challenging to perform large-scale RNA-seq based transcriptomic studies. Although we are working on the first draft of Nile grass rat genome, a well annotated genome typically requires several rounds of manually reviewing curated transcripts and can take years to achieve. To solve this problem, we developed a Comparative RNA-Seq Pipeline (CRSP), integrating a comparative species strategy but not depending on a specific sequenced genome or species-matched reference transcripts. Benchmarking suggests the CRSP tool can achieve high accuracy to quantify gene expression levels. In this study, we generated the first ultra-deep (2.3 billion × 2 paired-end) Nile grass rat RNA-seq data from 59 biopsy samples representing 22 major organs, providing a unique resource and spatial gene expression reference for using Nile grass rat as a model to study human diseases. To facilitate a general use of CRSP, we also characterized the number of RNA-seq reads required for accurate estimation via simulation studies. CRSP and documents are available at: https://github.com/pjiang1105/CRSP . Highlights CRSP is a novel software tool which can quantify gene expression levels from RNA-seq data for species lacking both a sequenced genome and corresponding reference transcripts. Nile grass rat is a unique diurnal rodent species (day active but not night active) with several unique characteristics making it as a useful model to study diet-induced type 2 diabetes and other physiological or behavioral processes due to its diurnal nature. We generated the first ultra-deep (2.3 billion × 2 paired-end reads) Nile grass rat RNA-seq data from 59 biopsy samples representing 22 major organs, providing a unique resource and spatial transcriptomic reference (e.g., tissue gene expression baseline) for using Nile grass rat as a model to study human diseases.