pseudoknot in radial layout?

[ceheitsch]

How does the radial layout tool handle peudoknots?

[maxieds]

@ceheitsch
I have tried my best to answer the referee's question in this WIKI subsection. Please do check me on the analysis I used. Also, let me know if this is detailed enough to address the question you had at hand.

Unfortunately, the structure you initially suggested that some group members are working with for other projects (RMRP.dbn) proved to be a little too complicated to be useful in answering the question. I found a simpler example on RFam (Hepatitis A virus substructure / knot). This is the example I defer to in the WIKI. Basically, my argument from visual inspection (many algorithmic links also suggested) boils down to the next comparison:

You can reproduce my results with

RNAfold -v hepatitis-rfam.fasta 
Processing 1. input file "hepatitis-rfam.fasta"
UUAAACAAAUUUUCUUAAAAUUUCUGAGGUUUGUUUAUUUCUUUUAU-CAGUAAAU
.((((((((((((............))))))))))))................... ( -6.60)

and/or using the resulting DBN file to load into RNAStructViz:
hepatitis-rfam.dbn.txt

[maxieds]

@ceheitsch @afafbioinfo
Looking back at what I added to the WIKI last week, I have another question, which I will ask the biologists in the room to keep me honest :)

It seems that the nice example of the pseudoknot I used from the RFAM database depends very much on the sample, which in the example I used the ViennaRNA utility (command line tool) RNAfold to generate on-the-fly. The algorithm we are using to plot the radial layouts is simple as far as this is concerned. Is one possible facet to address here in answering the referee question that we only have visual plots for the radial layouts where the secondary structure is completely specified at runtime? Note again that RNAStructViz does not allow users to input their samples in FASTA format alone.

So, if the pseudoknot affects how samples that are meaningful get generated, but we need actual explicit sample data to generate the plots on our end, this seems to be a moot point? In other words, I am not yet convinced that RNAStructViz does anything special with the pseudoknots other than plot a secondary structure that was helped to be determined by its presence.

I hope that question makes sense at this hour.