Thank you for your help indeed! Renaming them would have never crossed my mind!

Best regards,

Helena

On 15.1.2016, at 22.27, Oleg Sobolev <[email protected]> wrote:

Hi Helena,

I just spotted that residue 249 is close to 252. To resolve this, you will have to rename conformations of residue 249 (exchange A and B altlocs), assing all partial occupancies to 0.5 to make things simple and add it to selections:

selection = chain A and (resseq 252 or resseq 249) and altloc A
selection = chain A and (resseq 252 or resseq 249) and altloc B or chain S and resseq 1702 and altloc B

Renaming is necessary because conformaion B of residue 249 is close to conformation B of residue 252.

Best regards,
Oleg Sobolev.

On Fri, Jan 15, 2016 at 12:13 PM, Oleg Sobolev <[email protected]> wrote:
Hi Helena,

The problem was in altlocs, occupancies and selections.

First, simple case of 249 residue. I don't see why residue 249 is relevant to 252 or to water 1702 because it is rather far from them. Therefore it should be treated separately. Residue 249 simply has 2 alternative conformations of side chain (including CA atom). We set occupancy of A and B conformations to 0.5 (that is in your file). Also we need to set occupancy to 1 for N, C and O atoms in this residue to prevent their occupancy from refinement (in your file they are also present with 0.5 occupancy).

The second case: coupled conformations of 252 and water 1702.
For partially occupied water molecules you also need to define appropriate altloc. In your case it is present only when the residue 252 in B conformation, therefore water molecule also should have B altloc and partial occupancy (the same as B conformation of residue 252, 0.49). The residue 252 is split to alternative conformations completely, and this is OK. Next, to let phenix.refine know that these conformations are coupled, you need the following parameter file:
refinement {
  refine {
    occupancies {
      constrained_group {
        selection = chain A and resseq 252 and altloc A
        selection = chain A and resseq 252 and altloc B or \
          chain S and resseq 1702 and altloc B
   }
  }
 }
}
or enter these selections in appropriate place in GUI (in one constraint occupancy group).

Here each selection will get the same occupancy for all atoms in it, therefore we have both 252 residue and water residue and use altloc B. We don't have water with altloc A (but you may consider it) therefore the first selection contains only residue 252. These two selections present in one constrained group ensures that occupancies of these selections will add up to 1.

You can do all the model manipulations in Coot. 


I'll send you corrected pdb file shortly off-list.

The useful reading on topic would be

"13 typical occupancy refinement scenarios and available options in phenix.refine" by Pavel Afonine in the latest Computational Crystallography Newsletter:


Case #7 is almost your case, the only difference is one of your residues is water and it has only one conformation.

Best regards,
Oleg Sobolev.

On Wed, Jan 13, 2016 at 11:59 PM, Helena Taberman <[email protected]> wrote:

Hi,

Thanks for your advice. It’s still not working. I use the phenix gui version and if I set, for example,

chain A and resid 249
chain A and resid 252
chain S and resid 1702

as a constrained group at the occupancy refinement selections (the two residues have 2 conformations both and the water is there just if the conformation of residue 252 is B) the refinement just makes the residues a constrained group alone making even their main chain occ. something like 0.49 as for the both of the side chains too. And it also kicks them a bit out of their densities. So how to continue? What am I doing wrong?

Best regards,
Helena

Folmer Fredslund <[email protected]> kirjoitti 11.1.2016 kello 15.20:

Hi Helena,

I believer this is the relevant bit from the manual:
https://www.phenix-online.org/documentation/reference/refinement.html#occupancy-refinement (in the examples)

  1. Complex occupancy refinement strategy (combination of various available occupancy refinement types):

    % phenix.refine data.hkl model.pdb strategy=occupancies occ.params

The amount of atom selections makes it inconvenient to type them all from the command line. This is why the parameter file occ.params is used and it contains following lines:

refinement {
  refine {
    occupancies {
      individual = element BR or water
      individual = element Zn
      constrained_group {
        selection = chain A and resseq 1
      }
      constrained_group {
        selection = chain A and resseq 2
        selection = chain A and resseq 3
      }
      constrained_group {
        selection = chain X and resname MAN
        selection = chain X and resseq 42
        selection = chain X and resseq 121
      }
      remove_selection = chain B and resseq 1 and name O
      remove_selection = chain B and resseq 3 and name O
    }
  }
}


Hope this helps,
Folmer

2016-01-11 13:14 GMT+01:00 Christian Roth <[email protected]>:

Hi Helena,
Indeed you have to group the relevant aa and the respective water together in a constrained group to avoid your problem and get the occupancies add up to 1. This us described in the manual under occupancy refinement. I can't access the manual at the moment to give you a more detailed answer.

Cheers

Christian

On 11 Jan 2016 11:56, "Helena Taberman" <[email protected]> wrote:
Hi,

I’m trying to refine a structure determined to 1.6 Å resolution. There’s several residues with additional conformations and waters with partial occupancies stating that they would be there when the conformation of the residues allows it. When I try to refine these conformations and partially occupied waters together the refinement kicks the second conformations and water molecules out of their densities, since they are obviously too close to the other additional conformations of the residues or the following residues. What could I do? Should I somehow group them together or what?

Best regards,
Helena

Helena Taberman, Ph.D.

University of Eastern Finland
Department of Chemistry
Joensuu Campus
P.O. Box 111
FI-80101 Joensuu, FINLAND

Mobile: +358 50 337 2488



_______________________________________________
phenixbb mailing list
[email protected]
http://phenix-online.org/mailman/listinfo/phenixbb
Unsubscribe: [email protected]

_______________________________________________
phenixbb mailing list
[email protected]
http://phenix-online.org/mailman/listinfo/phenixbb
Unsubscribe: [email protected]



--
Folmer Fredslund


_______________________________________________
phenixbb mailing list
[email protected]
http://phenix-online.org/mailman/listinfo/phenixbb
Unsubscribe: [email protected]