Researchers from China report on results of a computational analysis on gene expression profiles from a wide range of mouse cell and tissue types. Using RNA sequencing data on 921 samples representing almost 300 mouse cell or tissue types, the team came up with computational approaches for predicting germ layers and cell type identity based on expression profiles. In the mouse, for example, the approach revealed eight cell type domains, beyond the simple distinctions already known to exist between ectoderm, mesoderm, and endoderm germ layers. "This model has potential implications for understanding development, adult tissue homeostasis, and manipulation of cell types in vitro," the authors write.
A team from the Netherlands describes a web-based tool for digging into human brain transcriptome data collections generated by the Allen Institute. The interactive web resource, known as BrainScope, is designed for visualizing adult and developing brain transcriptomes, using a computational approach known as t-distributed stochastic neighborhood embedding (t-SNE) to explore spatial transcriptome clues, gene co-expression modules, and gene functions that are specific to a given brain region. The researchers demonstrated the usefulness of the approach by applying it to adult brain transcriptome data from Allen Brain Atlas.
Thomas Jefferson University researchers present a strategy for sequencing mature transfer RNAs (tRNAs). The "Y-shaped adapter-ligated mature tRNA sequencing," or YAMAT-seq, method involves an adapter ligation step helped along by bacteriophage-mediated nick-ligation step that targets specific RNA structures, the team says, particularly regions of double-stranded RNA or RNA-DNA hybrid molecules. Based on their results, the authors argue that YAMAT-seq "provides [a] high-throughput technique for identifying tRNA profiles and their regulations in various transcriptomes, which could play important regulatory roles in translation and other biological processes."