Because these methods rely on sequence alignment, they are limited to sequences with high sequence similarity. While these methods are fast and deliver accurate results, they also require that the input sequences have been aligned correctly.
However, these methods are still too slow to perform an all-against-all scan of all windows from genomic sequences.Įxamples of the align-then-fold algorithms are RNAz, evofold and PETfold. To circumvent this problem, restricted, approximated or alternative versions of the Sankoff-style algorithm, including FOLDALIGN, Dynalign, CMfinder, LocARNA, Murlet, RAF and DAFS, have been published with/without application to ncRNA discovery. This algorithm, however, requires too high computational cost in terms of run-time and memory usage for any practical use. The most accurate methodology is the last one, and its pioneering work based on dynamic programming is presented by Sankoff. Comparative structure prediction can be classified into three groups, namely: align-then-fold, fold-then-align and simultaneous align-and-fold approaches. Ĭomparative methods, which in this case focus on structural conservation, are significantly more accurate than methods based on single sequences.
This structural conservation can be used to computationally locate ncRNA genes in the genomes. A significant part of the ncRNAs have a secondary structure, which is conserved between species. Furthermore, not all ncRNAs are expressed in all cell types, which also complicates detection of novel ncRNAs with experimental methods. While RNA-sequencing (RNA-seq) is routinely used to locate ncRNA transcripts, computational methods for detecting ncRNAs are needed since some ncRNAs are so lowly expressed that they can be hard if not impossible to detect in RNA-seq data without prior knowledge. In the current version of the human genome (Ensembl 86.38, June 2016), there are 22,219 non-coding genes and 20,441 protein coding genes annotated, indicating that ncRNAs are of great importance. Non-coding RNAs (ncRNAs), which are RNAs not translated into proteins, have many different functions within the cells.