User:ElNando888/Static Cloud Labs/CL17
ElNando888: Nascarnut, Hyphema and I were discussing earlier an interesting point about this lab, namely that practically none of the synthesized designs had the hairpin located at 43-52 formed correctly. While discussions were turning around intervening in the multiloop, or mis-using the 45-50 mismatch (by making it a real pair, like GU), I went another path.
I went to search the SCOR database (a real pity that this project is abandoned now more than 8 years) to see what kind of hexaloops would be very stable. Among the ones present in naturally occuring RNAs and not presenting a danger of creating a real pair at the mismatching points, the GUAACA sequence caught my attention.
Below, a comparison of the 3D predictions by RNAComposer between the sequence used by one of the top scorers of round 1, and this sequence.
Not sure what you guys think, but it seems to me that the second version is much more "orderly" (the sheared G/A noncanonical pair is formed, the bases keep stacked in the alignment of the underlying GC pairs) than the all-A one, which I would think, may mean that it would be more stable. But will that be enough to hold the stack below ?
Update : I checked the SCOR database again, there seems to be a bit of "conservation", like this would be a "motif", as sequences like
very interesting nando. Most times when you try to teach us something i usually just understand your concepts but this 3d rendition really shows off the 'orderly' fashion of this hexaloop sequence. I did submit a lab design for CL17 called New Idea that uses this sequence. lets see what happens. Hyphema
It seems to me that this hexaloop actually worked quite ok in your design, Hyphema. Unfortunate that other things didn't go as planned, but I'm taking this SHAPE data as a confirmation that this sequence is a nicely stable one, and I'm likely to reuse it.