Although a considerable portion of eukaryotic genomes is transcribed as long noncoding RNAs (lncRNAs), the vast majority are functionally uncharacterised. The rapidly expanding catalogue of mechanistically investigated lncRNAs has provided evidence for distinct functional subclasses, which are now ripe for exploitation as a general model to predict functions for uncharacterised lncRNAs. By utilising publicly-available genome-wide datasets and computational methods, researchers from the Garvan Institute of Medical Research present several developed and emerging in silico approaches to characterise and predict the functions of lncRNAs. They propose that the application of these techniques provides valuable functional and mechanistic insight into lncRNAs, and is a crucial step for informing subsequent functional studies.
- lncRNAs represent a large proportion of the transcriptome that is currently sparsely annotated.
- Expression-based experiments often yield a large number of lncRNAs cosegregating with the biological system being studied.
- The ability to effectively enrich candidate pools for lncRNAs most likely to be involved in the phenotype under study is crucial.
- Powerful computational methods for investigating lncRNA function and biology from experimental and sequence information are emerging.
- Combining several computational methods is an effective approach to maximise research findings and effectively deploy laboratory resources.
Summary of the Methods Presented
Each of the methods described in the main text probes a component of functional lncRNA biology. Those informative of the biological context are highlighted in light green, biological importance in dark green, regulatory target in dark blue, and functional mechanism in light blue. Using methods from each of these allows candidate lncRNA(s) to be identified, and further experiments targeting unknown biology to be designed.