Here’s an interesting tool I came across recently. It’s called RNAtips (RNA (temperature-induced perturbation of structure). It’s a web based tool that aims to tell you regions of RNA that are likely to undergo structural alterations due to temperature change. Here’s the citation:
The input for RNAtips is either one or two RNA sequences of the same length (in FASTA format). You then specify two temperatures to define the temperature range for which the RNA structural perturbation should be calculated (the default range is 32–39oC).
One of the types of output plots is at the top of this post, where the density of the most temperature-sensitive positions over the whole RNA sequence together with localization of clusters (red blocks) and localisation of nucleotide substitutions (if any) are plotted.
How does it do this? RNAtips deciphers those nucleotides within the RNA sequence, which change the most in their probability to form Watson–Crick bonds in response to a given temperature change. One of the first steps appears to be the calculation of probabilities of nucleotides being paired in a double-stranded conformation by using the RNAfold tool of the ViennaRNA package. The calculations performed are derived from those described in paper in RNA Biology in 2012. The web server output plots clusters of these temperature-sensitive positions within a sequence, therefore representing the most temperature-sensitive structural regions of that RNA.
This is a very interesting tool and it is useful to make interesting hypotheses relating to your RNA of interest. For example, how do the temperature sensitive regions of an RNA relate to possible alternative splice sites or predicted RNA binding site regions, or to particular secondary structure folds or sequence polymorphisms.
I wonder to what extent the temperature range selected when inputing the sequence data on the web-form influences things. The range 32-39oC is not really applicable to (most) plants. Does this mean the inferences made on RNA structure are ‘wrong’. It would be nice, also if the sequence plots (e.g. picture at the top of post) could be overlaid with multiple plots in order to compare multiple sequences.
You can follow ‘Its About Time’ on Twitter @Splice Time