Long non-coding RNAs found in human cells may exhibit more pronounced inter-individual variability in their expression than messenger RNA transcripts, according to a Genome Biology study by Austrian researchers. The team did RNA sequencing on human primary granulocyte cells from 10 healthy individuals, including seven individuals who were tested at several time points a month or more apart. In addition to uncovering lncRNAs not previously found in reference lncRNA sets, the investigators note that lncRNA expression varied from one individual to the next, surpassing the expression differences detected in mRNAs. Similar patterns turned up when they looked at data for other tissue types assessed by GTEx and GEUVADIS projects.
A University of Massachusetts-led team describes an effort to find and prioritize probably functional long non-coding RNAs in mammalian samples assessed by RNA sequencing. After developing and validating an automated filtering pipeline known as slncky, alongside another pipeline for aligning lncRNAs, which looks at features with potential evolutionary relevance, the researchers searched for signs of evolutionary selection in lncRNAs using data for vertebrates and other mammals. The search led to more than 200 intergenic lncRNAs under apparent selective constraint, they report, though most lncRNAs seemed to show species specificity.
Researchers in the UK and Australia introduce a computational approach for weeding out low-quality cells from single-cell RNA sequencing experiments. The pipeline is designed to gauge the quality of single-cell RNA-seq data using pre-processing, mapping, and quantification steps, in conjunction with a machine learning algorithm that takes a curated set of biological and technical features into consideration, the team notes. In their proof-of-principle experiments, the study's authors applied the pipeline to single-cell RNA-seq data for more than 5,000 T cells, bone marrow dendritic cells, or mouse embryonic stem cells, demonstrating that it compared favorably to existing methods for classifying low-quality cells.