[phenixbb] change sigma for an already restrained bond

[Pavel Afonine]

Hi Jorge,

great you got an answer from Nigel. I have not seen any reply on the 
phenixbb list so I replied as well. OK, it sounds like the problem is 
solved!

All the best!
Pavel

On 4/18/24 18:00, Jorge Iulek wrote:
> Dear Pavel,
>
>     Thank you; Nigel has also written me, I have tested according to 
> your directions and it works. One has to setup up cld=False as you 
> point (but this is not needed for angles, dihedrals).
>     Sincerely,
>
> Jorge
>
>
> On 4/18/24 21:35, Pavel Afonine wrote:
>> Hi Jorge,
>>
>> to answer your question and for the record, the following exercise:
>>
>> phenix.fetch_pdb 1yjp --mtz
>> phenix.refine 1yjp.{pdb,mtz} output.prefix=1 params.eff 
>> write_final_geo_file=true > & 1.zlog &
>> phenix.refine 1yjp.{pdb,mtz} output.prefix=2 params.eff 
>> write_final_geo_file=true cdl=false > & 2.zlog &
>>
>> confirms that you can override standard bond lengths and sigmas.
>>
>> In the first case I see (both geo files):
>>
>> User supplied restraints: 1
>> Sorted by residual:
>> bond pdb=" N   GLN A   5 "
>>      pdb=" CA  GLN A   5 "
>>   ideal  model  delta    sigma   weight residual
>>   1.456  1.459 -0.003 1.21e-02 6.83e+03 4.30e-02
>>
>> and
>>
>> User supplied restraints: 1
>> Sorted by residual:
>> bond pdb=" N   GLN A   5 "
>>      pdb=" CA  GLN A   5 "
>>   ideal  model  delta    sigma   weight residual
>>   1.458  1.458 -0.000 1.19e-02 7.06e+03 2.44e-04
>>
>> And in the second case I see
>>
>> User supplied restraints: 1
>> Sorted by residual:
>> bond pdb=" N   GLN A   5 "
>>      pdb=" CA  GLN A   5 "
>>   ideal  model  delta    sigma   weight residual
>>   2.000  1.459  0.541 1.00e-02 1.00e+04 2.93e+03
>>
>> and
>>
>> User supplied restraints: 1
>> Sorted by residual:
>> bond pdb=" N   GLN A   5 "
>>      pdb=" CA  GLN A   5 "
>>   ideal  model  delta    sigma   weight residual
>>   2.000  1.987  0.013 1.00e-02 1.00e+04 1.79e+00
>>
>> NOTE the nuance: you need to disable CDL for this to work as you want.
>>
>> Pavel
>>
>> On 4/16/24 11:23, Jorge Iulek wrote:
>>> Hi,
>>>
>>>     Well, people are very busy, but maybe one starting information: 
>>> is tightening of one single bond (through its sigma value) meant to 
>>> be possible?
>>>     Thanks!
>>>
>>> Jorge
>>>
>>> On 4/12/24 16:45, Jorge Iulek wrote:
>>>> Dear all,
>>>>
>>>>      I would like to change the sigma value for an already 
>>>> restrained  bond.
>>>>      I enter for phenix.refine:
>>>> "
>>>>      bond {
>>>>        action = add delete *change
>>>>        atom_selection_1 = chain D and resid 493 and name CA
>>>>        atom_selection_2 = chain D and resid 493 and name CB
>>>>        symmetry_operation = None
>>>>        distance_ideal = 1.532
>>>>        sigma = 1.02e-2
>>>>        slack = None
>>>>        limit = -1.0
>>>>        top_out = False
>>>>      }
>>>>
>>>> "
>>>>
>>>>      In the log I get:
>>>>
>>>> "
>>>> ...
>>>>    Custom bonds:
>>>>      bond:
>>>>        atom 1: "ATOM  14355  CA  ALA D 493 .*.     C  "
>>>>        atom 2: "ATOM  14358  CB  ALA D 493 .*.     C  "
>>>>        symmetry operation: x,y,z
>>>>        distance_model:   1.511
>>>>        distance_ideal:   1.532
>>>>        ideal - model:    0.021
>>>>        slack:            0.000
>>>>        delta_slack:      0.021
>>>>        sigma:            0.0102
>>>>      Total number of added/changed bonds: 1
>>>> ...
>>>> "
>>>>
>>>>      But in the .geo file I read:
>>>>
>>>> "
>>>> User supplied restraints: 1
>>>> Sorted by residual:
>>>> bond pdb=" CA  ALA D 493 "
>>>>       pdb=" CB  ALA D 493 "
>>>>    ideal  model  delta    sigma   weight residual
>>>>    1.532  1.511  0.022 1.28e-02 6.10e+03 2.85e+00
>>>> "
>>>>      1.28e-02 is the default value.
>>>>
>>>>      On the other hand, sigma changes for angles and dihedrals seem 
>>>> to work as expected (the intended values are echoed to the .geo file).
>>>>      To add up, if a bond is *added*, the value is in the .geo file.
>>>>
>>>>      I observe this for both
>>>> 1.20.1-4487
>>>> 1.21-5207
>>>>      versions.
>>>>
>>>>      So, in short, should I be allowed to *change the sigma* for a 
>>>> specific *bond*? How should I do it?
>>>>      Thanks
>>>>
>>>> Jorge
>>> _______________________________________________
>>> phenixbb mailing list
>>> phenixbb at phenix-online.org
>>> http://phenix-online.org/mailman/listinfo/phenixbb
>>> Unsubscribe: phenixbb-leave at phenix-online.org
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://phenix-online.org/pipermail/phenixbb/attachments/20240418/4b19a51a/attachment.htm>