Hi Edward,
I recently came across this belt of disordered lipid/detergent around a membrane protein. It was particularly prominent in the low angle reflections (say 15-20A and lower resolution). Specifically, from non density modified experimental maps, the boundary between the detergent head groups and water was well defined at low resolution creating a "shell" of density.
Are there any models that account for this scattering in refinement? Most refinement programs I've tried simply reject these reflections outright despite being reasonably measured (from merging statistics, and reflections were nowhere near the beamstop or overloaded).
Whereas simply cutting the resolution may be suitable for models that are complete, I've found that it complicates the interpretation of the electron density of the parts of the protein close to the detergent layer (for example, solvent accessible loops or N/C termini or domains) in early stages of model building.
F
On Sep 28, 2015, at 9:59 AM, Edward A. Berry <[log in to unmask]> wrote:
> Is it a membrane protein? Membrane proteins often co-purify with
> a belt of disordered lipid/detergent around the part of the protein
> that was in the hydrophobic phase of the membrane. This gets counted
> as "solvent" and as a result I think membrane proteins have on the
> average a higher solvent content than soluble proteins.
> eab
>
> On 09/28/2015 04:42 AM, Kavyashree Manjunath wrote:
>> Dear All,
>>
>> Thank you all for your suggestions. There is a huge
>> solvent cavity inside the crystal. And there are no
>> additional densities in this region to fit a monomer.
>> So as all of you suggested, probably this is an outlier
>> and PHASER solution in right.
>>
>> Thank you
>> Regards
>> Kavya
>>
>>> Hi Kavya,
>>>
>>> The important point here is that the calculator provides 'probabilities'
>>> and 'estimates' based on the most probable values observed in the PDB. If
>>> your protein deviates from a number of simple assumptions and 'average
>>> properties' then the predictor must invariably but statistically correct
>>> assign a low probability for your monomer or assembly.
>>>
>>> In the original calculator,
>>> http://www.ruppweb.org/mattprob/default.html
>>> the output is more verbose and following a suggestion of Dale Tronrud,
>>> emphasizes probability and lists more digits for the probability as not to
>>> confuse people by stating 0.00 for very low probabilities. Maybe this can
>>> be adapted in the Phaser output.
>>>
>>> More discussion of the limits of such predictions can be found in the
>>> kernel update paper:
>>> http://scripts.iucr.org/cgi-bin/paper?S1399004714005550
>>>
>>> Best, BR
>>>
>>>
>>> On Mon, Sep 28, 2015 at 12:25 AM, Randy J. Read <[log in to unmask]> wrote:
>>>
>>>> Hi,
>>>>
>>>> I don't think you would get such low R-factors so readily if you were
>>>> missing half the structure. 73% solvent is unlikely but not impossible.
>>>> If
>>>> you look at the packing, is there a connected lattice that could form a
>>>> plausible crystal? If your protein is a dimer, is there an appropriate
>>>> dimer generates by symmetry?
>>>>
>>>> Another possibility is that there is statistical disorder, if you don't
>>>> see plausible packing.
>>>>
>>>> Best wishes,
>>>>
>>>> Randy Read
>>>>
>>>> ----
>>>> Randy J. Read
>>>>
>>>>> On 28 Sep 2015, at 07:54, Kavyashree Manjunath
>>>> <[log in to unmask]>
>>>> wrote:
>>>>>
>>>>> Dear users,
>>>>>
>>>>> I am working on a 25kDa protein. The data was processed in
>>>>> P6122 space group. According to the Matthews coefficient
>>>>> calculated by "cell content analysis" program in CCP4, it is
>>>>> a dimer with a probability of 0.98.
>>>>>
>>>>> -----------------------------------------------------------
>>>>> Nmol/asym Matthews Coeff %solvent P(3.53) P(tot)
>>>>> _____________________________________________________________
>>>>> 1 4.63 73.44 0.02 0.01
>>>>> 2 2.31 46.88 0.97 0.98
>>>>> 3 1.54 20.32 0.00 0.00
>>>>> -----------------------------------------------------------
>>>>> I searched for dimer, but Phaser did not give any solution.
>>>>> but gave a solution when searched for a monomer.
>>>>>
>>>>> One round of refinement with the monomer, resulted in an R
>>>>> and Rfree of 0.26 and 0.32 respectively.
>>>>>
>>>>> Why PHASER is not giving a dimer? What problem could the data
>>>>> have in such cases?
>>>>>
>>>>> Thank you
>>>>> Regards
>>>>> Kavya
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>> --
>>>>> This message has been scanned for viruses and
>>>>> dangerous content by MailScanner, and is
>>>>> believed to be clean.
>>>>
>>>
>>>
>>>
>>> --
>>> -----------------------------------------------------------------
>>> Bernhard Rupp (Hofkristallrat a. D)
>>> 001 (925) 209-7429
>>> +43 (676) 571-0536
>>> [log in to unmask]
>>> [log in to unmask]
>>> http://www.ruppweb.org/
>>> -----------------------------------------------------------------
>>> The hard part about playing chicken
>>> is to know when to flinch
>>> -----------------------------------------------------------------
>>>
>>> --
>>> This message has been scanned for viruses and
>>> dangerous content by MailScanner, and is
>>> believed to be clean.
>>>
>>>
>>
>>
>>
>> --
>> This message has been scanned for viruses and
>> dangerous content by MailScanner, and is
>> believed to be clean.
>>
|