Many of these proteins are intrinsically disordered proteins (IDPs), meaning they don't actually have a set 3D structure, but instead existing in a kind of 'cloud' (ensemble) of interconverting conformations. A consequence of this is that there is no 'one' representative structure, due to this conformational heterogeneity.

People used to (very reasonably) assume that 'disordered' just meant totally random, so the thought was these disordered proteins behaved like a random polymer. However, just like in folded proteins, the amino acid sequence of these IDPs has a major impact on the way these clouds of conformations behave. Happy to discuss more here or 'offline' (see my profile for contact info) - this is basically my whole PhD, so, you know, I can go more in depth...

The sequences have only recently been found, and the interesting proteins don't share much homology.

Check out http://kumamushi.org for a lot if technical detail in the genome. jfarlow in comments below points to one of the interesting proteins that apparently provides dna protection.