[phenixbb] Occupancy refinement issue

[Pavel Afonine]

Hi Andrzej,

what you are trying to do is likely possible, we just need to explain it 
right to phenix.refine.
I can help you if you send me the PDB file or its fraction containing 
relevant atoms, and indicate occupancies of which atoms should be 
refined and how they should be constrained.

Pavel

On 10/7/11 1:51 AM, AFL wrote:
> Dear All,
>
>  I would like to refine the occupancy of _all_ components together for 
> the following system:
> - free LYS (chain A and resname LYS)
> - PLP-LYS complex (chain A and resname LLP)
> - PLP-GLU complex (chian D and resname PPE)
>
>  The LYS and LLP are defined as alternative conformations.
>
>  What I've tried so far is:
> - just phenix.refine without any additional constraints and the result 
> is occupancy refinement of LYS vs. LLP and separately refined PPE.
> - constrained_group {
>     selection = chain A and resseq 90 # position of LYS, LLP
>     selection = chain C and resseq 1  # PPE
>   }
>   Result is occupancy refinement of LYS+LLP vs. PPE and total mess in 
> the active site (=ligands moved to new positions far away from where 
> they should be).
> - constrained_group {
>     selection = chain A and resname LYS # ALYS, BLLP doesn't work
>     selection = chain A and resname LLP
>     selection = chain C and resname PPE
>   }
>  Results in occupancy refinement of LYS vs. PPE and the LLP occupancy 
> is set to 0.00.
>
>  So, the big question is:
> - is what I'm trying to do doable at all?
> - what would be the right way to do it?
>
>  I will appreciate all and any suggestions on how to proceed.
>
> Regards, Andrzej Lyskowski
>