[phenixbb] ADP refinement selection

Pavel Afonine pafonine at lbl.gov
Fri Jul 15 12:34:28 PDT 2011


Hi Hena,

1) refining group B-factor (one refinable B per group of atoms), 
phenix.refine does not change the mutual distribution of B-factors 
within the group. For example, if the original B-factors in the group were

10,20,30,40

after the group B-factor refinement they can be

10+b,20+b,30+b,40+b

where b is refinable parameter that can be any number that phenix.refine 
comes up with.

2) If you really want the B-factor to be all the same, then you need to 
reset them first to some meaningful number, say Wilson B or average B:

phenix.pdbtools model.pdb set_b_iso=50

then do refinement.

3) Note, that the default is

group_adp_refinement_mode = *one_adp_group_per_residue \
                                   two_adp_groups_per_residue 
group_selection

and if you change it to

group_adp_refinement_mode = one_adp_group_per_residue \
                                   *two_adp_groups_per_residue 
group_selection

phenix.refine will refine two B-factors per residue: one for side- and 
one for main-chain atoms of the residue.

4) For more details, see "On atomic displacement parameters..." article 
here:
http://www.phenix-online.org/newsletter/CCN_2010_07.pdf

5) In my experience highly restrained refinement of individual ADPs 
works almost always better than any of the group B refinement options 
mentioned above. phenix.refine uses unique type of restraints for 
individual isotropic B refinement that we introduced back in 2005: see 
above article for details.

Pavel.

On 7/15/11 12:17 PM, Hena Dutta wrote:
> Hi,
> I was trying to get one ADP per residue using the command as suggested 
> in the PHENIX document:
>
> *Refining group isotropic B-factors*
>
> One B-factor per residue:
>
> % phenix.refine data.hkl model.pdb strategy=group_adp
>
>
> See a part of my out pdb file (column 11):
>
> ATOM   6971  C2  ADE F  15     -29.960 -20.547 -31.167  1.00 23.82      F    C
>
> ATOM   6972  N1  ADE F  15     -30.382 -20.054 -32.336  1.00 24.20      F    N
> ATOM   6973  C6  ADE F  15     -30.077 -20.702 -33.471  1.00 21.52      F    C
> ATOM   6974  N6  ADE F  15     -30.496 -20.192 -34.629  1.00 20.08      F    N
>
> ATOM   6975  C5  ADE F  15     -29.321 -21.893 -33.361  1.00 21.08      F    C
> ATOM   6976  N7  ADE F  15     -28.823 -22.777 -34.307  1.00 22.85      F    N
> ATOM   6977  C8  ADE F  15     -28.193 -23.670 -33.593  1.00 21.80      F    C
>
> ATOM   6978  C2' ADE F  15     -27.648 -25.781 -31.357  1.00 22.11      F    C
>   
>
> But, I got different temperature factors for different atoms in the same residue. Is that what I should get? How can I create one
>
> temperature factor for each residue?
>
> Regards...
>
> Hena
>
>
> On Thu, May 26, 2011 at 11:05 AM, Pavel Afonine <pafonine at lbl.gov 
> <mailto:pafonine at lbl.gov>> wrote:
>
>     Hi Christian,
>
>     the problem is that you did not ask phenix.refine to refine the
>     rest of the atoms isotropically (with your command below you only
>     refine anisotropic ADPs of atoms in chain Q, and the B-factors of
>     all the other atoms are not refined):
>
>>          adp {
>>            individual {
>>              anisotropic = "chain Q"
>>            }
>
>     I would do it like this:
>
>     adp {
>       individual {
>          anisotropic = chain Q and element CL
>          isotropic = not (chain Q and element CL)
>       }
>     }
>
>     Pavel.
>
>
>     _______________________________________________
>     phenixbb mailing list
>     phenixbb at phenix-online.org <mailto:phenixbb at phenix-online.org>
>     http://phenix-online.org/mailman/listinfo/phenixbb
>
>
>
> _______________________________________________
> phenixbb mailing list
> phenixbb at phenix-online.org
> http://phenix-online.org/mailman/listinfo/phenixbb

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://phenix-online.org/pipermail/phenixbb/attachments/20110715/0cc58791/attachment.htm>


More information about the phenixbb mailing list