[phenixbb] following on the anomalous map

Fri Aug 29 03:05:26 PDT 2014

Xtriage assessment critically depends on the quality of the sigma's.
Suboptimal data processing can lead to thinking there is anomalous
signal when there isn't. The CC1/2 for the anomalous data could be
more informative.

The other possibility is of course that your solvent continuum is
anomalous: http://www.ncbi.nlm.nih.gov/pubmed/16714785

HTH
P

On 29 August 2014 02:24, Gabor Bunkoczi <gb360 at cam.ac.uk> wrote:
> Hi Charles,
>
> I second Nat's explanation on anomalous signal estimation (although in my
> experience, xtriage tends to err more often on the pessimistic side than the
> other way round). However, with an almost perfect model and data collected
> at the peak, the Br peaks should be enormous. Have you looked at the
> Phaser-EP logfile? What is the largest peak found in the map? (I can have a
> look, but please send this logfile to my private email address, and to the
> list.)
>
> Perhaps it would also help if you re-ran the Phaser-EP job, searching for Br
> as well as S atoms. With a good model, it is not uncommon that these could
> be located, even from short wavelength data. This would indicate that the
> anomalous signal you observe comes from sulfur atoms, and not from bromine,
> with the obvious conclusion that Br is either not present or disordered.
>
> BW, Gabor
>
>
> On 2014-08-28 14:59, CPMAS Chen wrote:
>>
>> HI, Gabor,
>>
>> Thanks for you suggestion.
>>
>> I tried this method, it did not give out NO sites for Br.
>>
>> why there is apparent anomalous signal and I cannot identify any Br
>> sites?
>>
>> As Nat pointed out, other atoms should have very weak anomalous signal
>> diffracted at Br absorption peak (3.8e- for Br, versus 0.2e- for the
>> sulfurs). In the difference map, I can clearly see some peaks at high
>> sigma (> 6), but the anomalous difference map(they are all generated
>> by phenix.refine or phenix.maps) is almost featureless, or the peaks
>> has no overlaps. The anomalous difference gives the possible place of
>> Br while the difference map gives the possible place of the whole
>> ligand, and I assume there should be some overlaps between them. Am I
>> right?
>>
>> Charles
>>
>> On Wed, Aug 27, 2014 at 6:28 AM, Gabor Bunkoczi <gb360 at cam.ac.uk>
>> wrote:
>>
>>> Hi Charles,
>>>
>>> I could be totally misunderstanding what you are trying to achieve,
>>> but I think what you need to do is an MR-SAD calculation. First,
>>> solve the structure with MR (use the macromolecule from the non-Br
>>> structure, or use your current model if you have done refinement on
>>> it), and use the Phaser-EP GUI, and select "SAD starting from MR
>>> model". This should highlight any Br you may have, and gives you a
>>> PDB-file with all the peaks (you can also get a proper anomalous
>>> map, but this is only useful for visualization - the peak search
>>> will identify any peaks you may have).
>>>
>>> Let me know if you need more details!
>>>
>>> BW, Gabor
>>>
>>> On 2014-08-26 14:49, CPMAS Chen wrote:
>>>
>>> Thanks. Nat.
>>>
>>> Since I have some MET and CYT in the protein, could I try to supply
>>> their sulfur as the initial anomalous scatters?
>>>
>>> As for the anomalous isomorphous difference map, would it be useful
>>> I
>>> compare the datasets acquired at 1A wavelength and at the Br
>>> absorption peak (~0.92A)?
>>>
>>> Appreciate your help
>>>
>>> Charles
>>>
>>> On Mon, Aug 25, 2014 at 5:08 PM, Nathaniel Echols <nechols at lbl.gov>
>>> wrote:
>>>
>>> On Sun, Aug 24, 2014 at 6:46 AM, CPMAS Chen <cpmasmit at gmail.com>
>>> wrote:
>>>
>>> Here is the point I am not clear. If I am using phenix.refine to
>>> generate LLG map, how do I pick the anomalous group since I have
>>> not placed them in the model yet?
>>>
>>> You can't. The LLG map only becomes really useful once you have
>>> some anomalous scatterers placed and refined - this is how Phaser
>>> substructure completion works. If your molecules have no other
>>> significant anomalous scatterers other than the expected Br, the
>>> LLG
>>> map won't do you much good.
>>>
>>> By the way, when I choose ion_placement and specify Br, the result
>>> comes with no Br.
>>>
>>> Not too surprising, since the code is tuned to look for ions, not
>>> part of a covalent molecule, and halides also tend to bond
>>> non-specifically and we haven't figured out how to deal with that
>>> yet.
>>>
>>> I want to find whether the Br-containing ligand is seen in my
>>> protein which I have a high resolution structure available.
>>>
>>> I have data collected at Br wavelength, peak or higher position.
>>> Phenix.xtriage reported that the anomalous signal is present to
>>> about 4A. However, both AutoSol or MR-SAD cannot identify the Br
>>> position. Simply say, AutoSol or MR-SAD can not generate any
>>> solution. Well, of course, the simple answer would be that there
>>> is no such ligand cocrystallized.
>>>
>>> The simplest explanation is that you just don't have enough
>>> anomalous signal to determine the substructure, which can be true
>>> even if your ligand is bound. Running experimental phasing to
>>> figure this out is unnecessary and time-consuming.
>>>
>>> Anyway, I am trying to see if the anomalous difference map or
>>> LLG(generated by phenix.maps, this would be the initial one I
>>> assume) can tell me anything more useful.
>>>
>>> So, my question on this topic would be what is a better way you
>>> guys would recommend to identify these Br-ligands? By the way, I
>>> did have the native datasets for the same protein with ligand.
>>>
>>> I think in this case I would start with the simple anomalous
>>> difference map. If you run phenix.find_peaks_holes (it's in the
>>> GUI, of course) and give anomalous data as input, it can pick out
>>> the highest peaks in the anomalous map. If the ligand really is
>>> bound and the Br site is ordered I would expect this to be
>>> detectable. Another alternative is to compute an anomalous
>>> isomorphous difference map between a dataset collected at or above
>>> (in eV) the Br peak, and a dataset collected below the peak. This
>>> will allow you to visualize the wavelength-dependent difference in
>>> anomalous scattering, and it's going to be specific for elements
>>> with absorption edges within that energy range. But I really don't
>>> think this should be necessary to answer your question.
>>>
>>> -Nat
>>
>>
>>  --
>>
>>  ***************************************************
>>
>>  Charles Chen
>>
>>  Research Associate
>>
>>  University of Pittsburgh School of Medicine
>>
>>  Department of Anesthesiology
>>
>>  ******************************************************
>>
>>  _______________________________________________
>>  phenixbb mailing list
>>  phenixbb at phenix-online.org
>>  http://phenix-online.org/mailman/listinfo/phenixbb [1]
>>
>>
>>  --
>>  ##################################################
>>
>>       Dr Gabor Bunkoczi
>>
>>       Cambridge Institute for Medical Research
>>       Wellcome Trust/MRC Building
>>       Addenbrooke's Hospital
>>       Hills Road
>>       Cambridge CB2 0XY
>>  ##################################################
>>
>> --
>>
>> ***************************************************
>>
>> Charles Chen
>>
>> Research Associate
>>
>> University of Pittsburgh School of Medicine
>>
>> Department of Anesthesiology
>>
>> ******************************************************
>>
>>
>>
>> Links:
>> ------
>> [1] http://phenix-online.org/mailman/listinfo/phenixbb
>
>
>
> _______________________________________________
> phenixbb mailing list
> phenixbb at phenix-online.org
> http://phenix-online.org/mailman/listinfo/phenixbb

-- 
-----------------------------------------------------------------
P.H. Zwart
Research Scientist
Berkeley Center for Structural Biology
Lawrence Berkeley National Laboratories
1 Cyclotron Road, Berkeley, CA-94703, USA
Cell: 510 289 9246
BCSB:      http://bcsb.als.lbl.gov
PHENIX:   http://www.phenix-online.org
SASTBX:  http://sastbx.als.lbl.gov
-----------------------------------------------------------------