Dear Tim,
Thank you for your detailed thoughts, which describe fairly clear the situation I meant. Instead of a 3D NOESY, I am working with solid-state restraint spectra, such as 2D DARR or 2D PAR where you can find either the TOCSY situation you described or overlaps as you depict for the NOESY.
Surely, within a single spectrum the frequency of a peak is very well defined at one place, but eventually not in others and analysis fixes the blob by putting the peak in the center of this blob, often rather far off from the "real frequency" and changes the (1) "standard deviation" a lot and should be fixed somehow (what you describe).
For me the combination of tools you described is very logical and helpful to me.
Thank you again for all you input
- Carolin
Ps: The idea to the tell the program that the frequency is like the isolated one comes actually from a tool used in CARA. I have no clue if that might help you in gaining ideas on peak shape fitting... but definately CARA is pretty good in that, too.
________________________________________
Von: CcpNmr software mailing list [[log in to unmask]]" im Auftrag von "Tim Stevens [[log in to unmask]]
Gesendet: Dienstag, 5. Juni 2012 16:41
An: [log in to unmask]
Betreff: Re: fast way to cahnge the assignments of an atom in different "pairs"
While keeping with the notion that you should always deal with real peaks
with real standard deviations, and this is how Analysis is mostly
constructed, there are occasional situations where you perhaps would want
to add peaks at averaged positions. For example in a TOCSY (where
intensity per se is not so important) it can be useful to have the
complete 'grid' of peaks for each spin system so you can see which
resonances underly big blobs and artefacts, so that you can see what has
or has not been accounted for. If this is done using the synthetic peak
list construction then there will be no undue influence on the chemical
shift average and the real peaks will be in a separate list.
Assigning real peaks to multiple resonances to record overlap will cope
with most situations, but there is a potential issue in that assigning
knowingly distorted peak centres is not representative of the chemical
shift average. However, in severe situations the merit value of a peak can
be set to zero (so it has no shift influence) or more generally the shift
weighting for the spectrum dimension can be adjusted, for example I often
reduce the influence of an indirect 3D NOESY 1H dimension because I don't
want intermediate assignments from ARIA to have an effect (some rarer
shifts can wander as the average changes and get associated with the wrong
peaks in later structure calculations).
Wayne: Maybe in the future we could do something better with peak shape?
T.
> Dear Brian, I completely agree and thank you for your opinion.
> I will play around with the tool to see what it does with overlaps.
> Thank you
> Carolin
>
> On Jun 5, 2012, at 2:12 PM, Brian Smith wrote:
>
>> Surely this is just wrong? Your crosspeaks represent individual
>> measurements each of which has associated noise & artefact. The (perhaps
>> weighted) average of the peak positions is what you want to use for any
>> downstream application such as CSI etc., shift matching peaks etc.. This
>> is is calculated and stored for you in the shift lists in analysis. Don't
>> waste time fiddling individual peak positions to sanitise (dare I say
>> falsify?) your data.
>>
>> Seuring Carolin <[log in to unmask]> wrote:
>>
>> Dear Wayne and Tim,
>>
>> this is pretty much what I was looking for! If I right-click on selected
>> peaks and then go to Assign:Unite Resonance the program aligns them (in
>> its way).
>> In crowded regions (overlap), I would like to influence the average
>> position myself - can I?
>> For example : Is there away to put this "average value" in a certain list?
>> Also, now I selected the resonances in the spectrum - is there a way to
>> choose the shifts to be aligned from a peaklist and then align them? Often
>> I have 10 peaks of a single residue, e.g. ProCa.
>>
>> What does the RMM in RMM:Assign:Unite Resonance stand for? I can't find
>> Assign in one of the resonance options.
>>
>> Thank you already very much!!! Your solution makes it already a lot easier
>> than before.
>> - Carolin
>>
>>
>> On Jun 5, 2012, at 11:49 AM, Wayne Boucher wrote:
>>
>>> Hello,
>>>
>>> Tim says:
>>>
>>> "What Carolin needs is probably RMM:Assign:Unite Resonance Positions
>> over the most isolated/representative peak, noting that this will affect
>> all the assigned resonances in the spectrum, i.e. so working at the Ha,Hb
>> peak will mean that these two resonance positions are spot on and all the
>> other peaks assigned to Ha or Hb in the spectrum will be moved to align
>> perfectly."
>>>
>>> Wayne
>>>
>>> On Mon, 4 Jun 2012, Seuring Carolin wrote:
>>>
>>>> Hi there,
>>>>
>>>> I cannot find a solution for how I can change all frequencies of an
>> atom in 1 spectrum at once ? In my spectra, 1 atom is often assigned in
>> several different peak pairs, e.g. LysCa is found in (1) LysCa-LysCa, (2)
>> LysCb-LysCa, (3) LysCa-LysCg... etc. LysCa does not have the exact same
>> frequency in all of them, mostly because of overlap or small intensity
>> values if I peak it automatically (Shift+Ctrl).
>>>> I can change the LysCa frequency by going through the whole peaklist
>> and edit the frequencies manually.
>>>> Is there a different, faster way?
>>>>
>>>> I very much appreciate your help!
>>>> Thank you !
>>>>
>>>> Carolin
>>>>
>
-------------------------------------------------------------------------------
Dr Tim Stevens Email: [log in to unmask]
Department of Biochemistry [log in to unmask]
University of Cambridge Phone: +44 1223 766018 (office)
80 Tennis Court Road +44 7816 338275 (mobile)
Cambridge CB2 1GA WWWeb: http://www.bio.cam.ac.uk/~tjs23
United Kingdom http://www.ccpn.ac.uk
-------------------------------------------------------------------------------
------ +NH3CH(CH(CH3)OH)C(O)NHCH(CH(CH3)CH2CH3)C(O)NHCH(CH2CH2SCH3)CO2- -------
-------------------------------------------------------------------------------
|