zlacker

> It ultimately relies on mapping sequence to structure.

So does every structural prediction method.

> if you give it a brand new fold with no relation to other folds, it cannot predict it

That will depend on the number of effective sequences, not the actual fold.

> I work in organisms that have virtually 0 sequence identity.

Then the problem is low sequence coverage, not the protein fold. On a side note, there are sensitive homology search protocols that rely very little on actual sequence identity.

>>flobos+(OP)
So then based on your counter arguments to the OP, have they mapped the entire protein universe ? Or should it say, the “already known protein universe” ?

>>bamboo+aU
Neither the protein sequence nor structure spaces have been fully explored, and the sequence set of UniProt does not represent every single extant protein. My answer is “no”.