The Significance of the Backbone in Eterna
The backbone has minimal significance for Eterna puzzles, where its only role is to "string together" the bases. That is, you can mutate any base that hasn't been "locked", but you can't actually move a given base to a different position in the puzzle.
In the lab, though, the situation is very different. The data we get back from the lab (the SHAPE data) comes from measuring the rate at which the SHAPE chemical (specifically 1-methyl-7-nitroisatoic anhydride, or 1M7) interacts with the backbone at each base position. The nucleic acid base itself (A, U, G or C) plays no role in the chemical reaction. So when we interpret the lab data, we are actually making inferences about the configuration of the backbone.
As a first approximation, the generally accepted explanation is that the rate at which the SHAPE chemical (or SHAPE probe, as it is often called) reacts at each base is an indication of how flexible the backbone is at that base. If the backbone is flexible enough, the thought goes, it can wiggle around to get into the right position to interact with the SHAPE probe; if the the backbone is being held rigidly, the SHAPE probe can not get access to the reaction site. Furthermore, there is an assumption in this simplified theory that unpaired bases will have "flexiible" backbones, while paired bases will have "rigid" ones.
But any Eterna players who have looked at lab data critically know that this can't be the whole story. Consider, for example, the SHAPE values for the closing base bair of a hairpin loop. It is quite common for the 3' base to be protected from the SHAPE probe (implying it is paired) while the 5' base is exposed to the SHAPE probe (impying it didn't pair). But pairing is a symmetric relation; A can't pair with B unless B pairs with A.
Backbone Units: Nucleotides versus Suites
Backbone Suite Conformers
To view in 3D, look "behind" the image until both sides appear to overlap, then relax your eyes to focus. If you are successful, you will see three copies of the image, with the one in the middle being the one that shows the depth effect. Note that for this to work, the distance on the screen between the two images has to be less than the distance between your eyes. If you can't get the middle images to merge, try using your browser to zoom out, making the images smaller and closer together.