An RNA-Seq atlas of gene expression in mouse and rat normal tissues

Abstract

Gene functionality is closely connected to its expression specificity across tissues and cell types. RNA-Seq is a powerful quantitative tool to explore genome wide expression. The aim of this study is to provide a comprehensive RNA-Seq dataset across the same 13 tissues for mouse and rat, two of the most relevant species for biomedical research. The dataset provides the transcriptome across tissues from three male C57BL6 mice and three male Han Wistar rats. We also describe our bioinformatics pipeline to process and technically validate the data. Principal component analysis shows that tissue samples from both species cluster similarly. We show by comparative genomics that many genes with high sequence identity with respect to their human orthologues also have a highly correlated tissue distribution profile and are in agreement with manually curated literature data for human. In summary, the present study provides a unique resource for comparative genomics and will facilitate the analysis of tissue specificity and cross-species conservation in higher organisms.

Chromosome-wide co-fluctuation of stochastic gene expression in mammalian cells

“…Hence, we predict that, the more negative the across-tissue expression correlation is between a protein complex member gene and H1 histones, the higher the likelihood that the gene is driven to be linked with other genes encoding members of the same protein complex. To verify the above prediction, we used a recently published RNA-seq dataset [52] to measure Pearson's correlation between the mRNA concentration of a gene that encodes a protein complex member and the mean mRNA concentration of all H1 histone genes across 13 mouse tissues. Indeed, the linked protein complex genes show more negative correlations than the unlinked protein complex genes (P = 0.012, one-tailed Mann-Whitney U test; Fig.…”

“…This analysis used the RNA-seq data from 13 mouse tissues [52] as well as the protein complex data aforementioned. We divided all protein complex genes into three groups: unlinked genes, linked genes, and evolved linked genes.…”

An information-theoretic approach for measuring the distance of organ tissue samples using their transcriptomic signatures

“…The number of available samples for each healthy human organ is provided in Table S1. For mouse and rat, we obtained RNAseq data from the database developed by Sollner et al (2017). For both species, the number of available samples per organ was equal to 3.…”

“…at various states (e.g., healthy, diseased, etc.) (Uhlen et al (2015(Uhlen et al ( , 2017; Yu et al (2015); Keen et al (2015); Mele et al (2015); Suntsova et al (2019); Sollner et al (2017)) using RNA-sequencing, a mature technology for quantifying gene transcripts in biological samples. A notable effort is the Human Protein Atlas (HPA) project (Uhlen et al (2015)), a Swedish-based program providing, among others, gene expression signatures for 37 healthy human organ tissues.…”

“…Such tools will not only help us understand the model capabilities and limitations but also reveal aspects we can improve in their design to optimize their physiological relevance and increase their value for precision medicine. Next-generation sequencing data (e.g., RNA-seq) is already utilized to determine the distance between organ tissue samples (Mele et al (2015); Suntsova et al (2019); Sollner et al (2017); Sudmant et al (2015)) in conjunction with classical mathematical tools such as the Euclidean distance, or dimensionality reduction techniques (e.g., Principal Component Analysis, Linear Discriminant Analysis, Uniform Manifold Approximation and Projection, etc.). However, all these methods exhibit significant limitations due to sensitivity to noisy measurements and outliers, inability to capture non-linear relations, etc.…”