On Thu, Jan 17, 2013 at 7:17 AM, Antony Oliver wrote:

I've noticed a slightly annoying "feature" when using a reference model to help in refinement of low-resolution structures. phenix.refine seems to take it's crystal parameters and space-group from whichever PDB was loaded last, even after setting one PDB as the "reference model" and the other as the "Input model".

Correct.

When running phenix.refine it gives you a nice error message telling you that the space group information is incorrect, and that you may have used the reference model's parameters by mistake… But surely, it should just (by default) only take (check?) the cell parameters and space group information from the "Input PDB"?

I agree that this is not well-designed - the problem however is that you can have multiple input PDB files, e.g. one for protein, one for ligand, etc. When a PDB file is added it isn't immediately obvious what its purpose is, but the assumption is that if it has symmetry, we want to use that information. It used to only pull information from the first file, but I received multiple complaints about this. I may be able to figure out something smarter, but it's not trivial. (One option would be to pop up a message asking if you want to use the symmetry, in the case where it is already defined from a previously loaded file, but I personally find pop-up messages annoying.) For now I will add a button that will let you choose which file to pull the symmetry from, so you can easily revert to the correct parameters. -Nat