I was recently running RosettaRemodel to build a disulfide between two residues. The first residue is a gly and the second is tyrosine. I ran into this error:
“Failed glycine check. Cysteines cannot replace glycine positions in the native pose.”
Is this because gly doesn’t have a CB atom so that the CB-CB distance cannot be measured? I can think of mutating the gly to other residues as a workaround. But is there any other way to build a disulfide bond involving a gly using RosettaRemodel?
I just checked the code for the error message in question. Quite simply, it just checks if two residues in question are glycines–so the answer to “is there any other way” is pretty much “no.” I imagine that the lack of a CB atom is the reason, and it stems from the prerequisites for the DisulfideMatchingPotential that RemodelDesignMover employs.
If you want to be adventurous, you could remove that check and futz around with DisulfideMatchingPotential, but I think far more concise is just to mutate your input structure to alanine (perhaps rename the residue from GLY to ALA in the PDB and then read it in using Rosetta; perhaps manually replace the alpha hydrogen with a CB at the appropriate coordinates–it probably won’t matter much because it’ll be getting rebuilt anyway).