Artificially made/optimized homologous proteins, have completely different folds and functions. Confirmed by NMR.

http://hwmaint.pnas.org/cgi/content/abstract/105/38/14412

95% identical DNA become an entirely different species too. This is why it is such a difficult problem that requires study. You can't just assume that using the same building blocks results in the same structure.

---

Rosetta Moderator: Mod.Sense

You can't just assume that using the same building blocks results in the same structure.

Hmm... I was not assuming anything. Just stated fact. Fact and publication that I think is really interesting one.

Why I found it amazing? Because at the moment whole field of protein structure prediction do use homologous sequences(70% similarity PDB model more then good enough) as a part of standard protocol, as a starting backbone for the fold. And that include Robetta or any other methods used in CASP.

And for a very good reason. In overwhelming majority of known cases, related sequences in vivo do produce very similar folds. Those proteins mentioned in the PNAS paper were artificially designed/optimized to archive such a strange property - almost complete homology while having different structures.

Anyway, I'd rather expect Rosetta project administrator to know that. As for banalities, 99% identical MS Excel formulas will produce completely different results. Almost identical social security numbers will belong to different people. Sky objects that indistinguishable to human eye could have completely different origins/mass/structure/distance.
Etc. and so on and so forth...