[phenixbb] constrained occupancy groups in phenix.refine

[Nathaniel Echols]

On Fri, Sep 3, 2010 at 8:23 AM, Lionel Costenaro <lcocri at ibmb.csic.es> wrote:
> occupancies {
>       constrained_group {
>         selection = "(chain A or chain B) and (resseq 105 or resseq 110:112
> or resseq 140:143) and altid A"
>
> After refinement, all alt_A selected residues do have the same occupancy,
> but not the same occupancies for alt_B and the sum for one residue (occu
> alt_A +occu alt_B) is not 1.
> Adding a second group with the same selection but "altid B", constrains also
> the alt_B occupancies but the sum is still not 1.

Do you mean you added a second "constrained_group" block?  This is
incorrect - you should use multiple selections instead:

occupancies {
      constrained_group {
        selection = "(chain A or chain B) and (resseq 105 or resseq
110:112 or resseq 140:143) and altid A"
        selection = "(chain A or chain B) and (resseq 105 or resseq
110:112 or resseq 140:143) and altid B"
      }
}

I think that will force the sum of occupancies to be 1.

-Nat