Grant: Discovery Grants Program - Individual

Small nucleolar RNAs (snoRNAs) are a large family of non-coding RNAs present in all eukaryotes and extensively characterized for their role in ribosome biogenesis. Over the past two decades several independent studies, mostly in human, have reported additional roles for small numbers of snoRNAs in gene expression regulation, but the extent of such functionality and its existence in other eukaryotes is unknown. Intriguingly, snoRNAs use diverse strategies for their expression in different eukaryotes, including intergenic vs intronic location and being encoded alone or as part of a snoRNA cluster, and snoRNAs can exist in multiple copies. The sets of annotated snoRNAs varies greatly across eukaryotes, from none to tens of thousands in genome annotations. Even in the best characterized organisms such as human, snoRNA annotations are problematic. The aim of this research programme is to determine and characterize snoRNAs across diverse eukaryotic organisms, enabling the characterization of their expression regulation, conserved function and evolution. The specific aims are to 1) Broaden and sharpen the snoRNA landscape in human. SnoRNAs are by far best characterized in human, but even in this organism, the annotations are still lacking. Our preliminary analyses indicate dozens of snoRNAs missing from annotations. We will sequence diverse tissues and cell types using our snoRNA-friendly sequencing pipeline to widen the range of organs and conditions that we consider, identifying missing snoRNAs and improving the classification of snoRNAs as expressed or pseudogenes. 2) Improve the annotations and the prediction of snoRNAs across eukaryotes. In addition to human, through collaborations, we have already applied our snoRNA-tailored pipeline to other eukaryotes and plan to sequence additional organisms to improve annotations in diverse eukaryotes, including the identification missing snoRNAs. These improved annotations will also be used as training sets to create accurate machine learning predictors of snoRNAs across eukaryotes, capable of distinguishing between expressed snoRNAs and pseudogenes, which will improve annotations even for organisms for which no sequencing has been carried out. 3) Redefine snoRNA families. The Rfam webserver is a widely used resource providing RNA family classification of structural RNAs, based on careful curation and covariance models. However, current Rfam snoRNA families were created more than 15 years ago and require updating to take into account new members in many eukaryotes. In collaboration with the Rfam team at the European Bioinformatics Institute, we will build and rebuild snoRNA homology models. We will also study snoRNA family expression regulation across eukaryotes. Mapping and characterizing snoRNAs in diverse eukaryotes through sequencing and computational prediction will considerably increase our understanding of the regulation, function and evolution of these ancient and central RNA regulators.