2 SHAPE questions are giving me the blues.

Good morning.

Eli suggested I post my question about the SHAPE data in the forum. If I am understanding correctly, the darker the blue the better the bond, and that “light blue” bases in loops are scored as yellow… so two questions:

In lab 3, I compared the top score top 97 to mine 93. They have less yellows in the stacks, and I understand that is why they get the extra points and such, but my bonds are much darker and more consistently so:

  1. it seems that the labs are scored in a simple blue/yellow paradigm. Does this mean that using non-continuious mode is the best way to view the results to see where your “score-lowering” areas are, (excepting the light blue in loops as light blue in stacks seems to be scored just as “blue”)? and

  2. if the darkness of the blue indicates the strength of the bonding, how is it that opposite sides of a bond are different shades of blue? Does it have to do with the electronegativity of the macromolecule such that a base on one side will be more reactive than the one on the other? Just a relatively uneducated and sleep-deprived guess.

Thanks in advance!

Anyway, hope something in there makes sense… i’ve tried to post pictures.

the top score in lab 3 (97)

my score in lab 3 (93)


It is best to look at the yellow/blue coloring in continuous mode. For the scoring system, we give the residue a point if: (1) it is supposed to be paired and it is less than 0.5, or (2) if it is supposed to be unpaired and it is greater than 0.25. [Here 0.5 corresponds to the boundary between yellow and blue.]

In terms of mismatches in color across a pair, it could be that the RNA actually misfolded into an alternative local structure there. Another alternative is that occasionally the chemical modifier 1M7 can react at paired residues at the ‘edge’ of a stack. We are tracking down this possibility now…

If you would also like to know which bases did not score a full point, they would most likely be:

Bases 30, 35, 41, 60, and 75 out of the 79 scored bases. The winner in comparison only had 2 bases that did not receive a full point, and those would be bases 30 and 41. As you can see, it had similar problems, the lonely base pair at 35-41 had issues with being detected in most designs, and I do not believe any designs gained the full 2 points there, and very few actually scored 1 point. As for base 30, I believe that falls under “the chemical modifier 1M7 can react at paired residues at the ‘edge’ of a stack”. In fact, this same thing has happened with many designs that have a single nucleotide bulge loop.

I think it’s all complicated by the fact that there are actually many shapes represented in solution at the same time and the scoring is based on a closest fit to the “average”.