Maybe! We tried to compare sequence vs. structure for all the genes in our favorite genome (C. auris) using FoldSeek to identify structural similarity. The scripts are here: github.com/maomlab/poc4 and the preprint is here: www.biorxiv.org/content/10.1...

Here's the relevant panel:

I saw the best bumper sticker today

We designed a new protein to fulfill the geometric constraints of the structure and showed that it too can complement the mutant

To determine whether these changes were impacting function, we were able to complement the thermotolerance defect of the C. auris mutant with the S. cerevisiae protein, despite only 20% identity

We determined that this protein was evolving fast, even faster than other proteins in the proteasome (like Pup2) or even its dimer protein Irc25. This was unexpected, as co-evolution would suggest that proteins involved in the same process would evolve at the same rate

A particularly interesting test case was an uncharacterized protein that looked like a proteasome assembly chaperone.

When we used a structural similarity search, we found a whole pile of proteins that had good structure hits but bad sequence hits