Challenges and promises of RNA diagnostics

Wednesday, October 14, 2015 | 3:00 to 4:00 p.m. ET

Thomas Tuschl, Ph.D.

Professor and Head of the Laboratory for RNA Molecular Biology

HHMI Investigator

The Rockefeller University

ttuschl@rockefeller.edu

Website

Summary

There is an increasing need in life and medical sciences to determine gene sequence and gene expression in cells and tissues including pathogenic organisms and viruses under normal and disease conditions as well as organismal development. The limitations in diagnostic and prognostic tools have also triggered interests in RNA present in extracellular fluid such as plasma and urine. RNA sequence analysis has the advantage of providing sequence as well as abundance information and contains information such as allelic sequence information, mutations, or translocations, traditionally only mined by genomic DNA sequencing. There is also reduced complexity at the RNA level as tissues or cells never express simultaneously all human genes. Furthermore, important abundant transcripts encoded by multicopy genes or accompanied by pseudogenes are difficult to evaluate by genomic DNA sequencing yet readily captured at the RNA level. The Tuschl laboratory has developed RNA sequencing (RNAseq) approaches over many years and has discovered the genes encoding miRNAs and piRNAs. The group has revisited all classes of RNAs, including mRNAs, in order to arrive at a cleanly annotated and minimally redundant human transcriptome including normal and disease-causing allelic variation, RNA editing sites, and isoforms. The emerging reference transcriptome is integrated in a large prototype database able to hold large experimental RNAseq data sets in order to supporting hierarchical annotation processes adaptable to any size of input RNA. This platform is able to generate in-depth reports of RNA abundance, specificity of expression across samples, detailed alignment reports and HTML displays, searching of reads and transcripts, and manually curating reference transcript entries. RNA reads without retrievable annotation are returned in form of assembled contigs to allow for discovery of novel genomic rearrangements, pathogens, or genes searching in public databases. Current software tools for alignment and mapping are fairly complicated and can only be used by domain experts while they still miss to explain a substantial portion of RNAseq data. Most recently, the Tuschl lab has developed automatable approaches for isolating extracellular RNA from plasma/serum and urine samples and studied the small RNA composition by RNAseq in nearly 1000 samples of normal subjects and patients including liver, heart, and kidney diseases. The experimental as well as analytical challenges will be reviewed.