The question was: "I am looking for a way to conserve side-chain conformations across mutate_replace. The default way muta_rep works is that it takes CHI1 and CHI2 angles from its database. However, if I have a molecular replacement solution and want to mutate the old structure's GLU to my structure's ASP, I would like O to take the CHI1 angle of my ASP from the previous GLU. Is there any way to do this ?"

You could keep a backup copy of your mol rep search model and try to use "merge_atoms" (don't know if it works if the residues are different, but assume it does); you probably need to do a "refi_zone" as well afterwards

Personally, I use the following 'protocol':

if the side chain dens is clear, then:

• - select the most promising rotamer
• - do move-zone on the residue; double-click the Calpha so it is the pivot point for rotations; move the side chain into the density
• - use tor_residue for long, floppy side chains

if the density is not clear:

• - if you have a similar molecule (e.g., Mol Rep search model), use lego_side_ch to select the rotamer that is most similar to the side chain of that molecule (provided it doesn't clash with anything else, of course)
• - if not, use any rotamer that might make sense
• - in regions where I'm not sure that I'm in register with the sequence, I keep a poly-Ala initially