Reply To: Non-ideal residue detected

[Anonymous]

As you’re probably aware, Rosetta runs on internal coordinates (2-body bond lengths, 3-body angles, and 4-body torsions.)

The physics of bond lengths and bond angles are very tight; there’s little variability. Almost all the action in folding a protein is changes in the bond torsions only.

So, to simplify, sometimes Rosetta runs on “idealized” proteins. The bond lengths and angles are set to perfect ideal coordinates, and then Rosetta can describe the protein structure with only the sequence and torsions, because the angles and bonds are already known. This is particularly useful for fragment insertion, where the fragments are represented by torsions only, assuming idealized angles/lengths.

I don’t know how to fix the error but that’s what it means. Check your inputs, maybe something needs to be run through idealize.cc first.