Re: R2 from R1rho & R1 -- March 27, 2014 - 09:26

Hi Edward.

Yep, I think some "porcelain functions" that will work on

relax base "plumbing" functions would be the best solution?

That could return info to screen, and return values in dictionarys, where the dictionary keys are also stored.

Actually, an extension of "calc_rotating_frame_params" in  "specific_analyses.relax_disp.disp_data"

which should take interpolated offset data as a possibility.

For 3D, I would prefer that we use gnuplot instead ?

I really really start liking the idea of the grace files.

Here you in "one" file carry around both the data and definitions to the plot.

For matplotlib, you would need to acquire that the user has relax and matplotlib installed at the same time.

That is not so good, and would be a problem for the compiled sources.

Unless you export python files, with the data inside, and then the plot commands to matplotlib.

Hm hm.

What is best?

Expect the user to know gnuplot , and have that installed?

Or that the user have matplotlib ?

I know that matplotlib on our RHEL system is only 0.99, and is not moving.

Gnuplot is old... And proven to work.

Reach for the future or backwards-compability?

Best

Troels

Troels Emtekær Linnet

2014-03-27 8:51 GMT+01:00 Edward d'Auvergne <edward@xxxxxxxxxxxxx>:

Hi Troels,

Storing such secondary dispersion parameters (Rex, R1rho'+Rex, etc.)
is not a good idea as they can be very difficult to keep up to date.
They are dependent on the base parameters of the model as well as the
experimental details (offsets, field strengths, etc.), so any change
in these values would require the secondary parameters to be updated -
but this approach is often a big source of future bugs.  It is better
to generate them when asked by a user function, as 'theta' and 'w_eff'
currently are.

A 3D graph as a surface would be quite interesting!  I have seen such
plots before, but I can't remember off the top of my head in which
papers that was.  For this we would need to add some basic
infrastructure to relax for plotting with matplotlib.  Grace does not
do 3D surface graphs.  Adding this infrastructure is rather trivial as
the code for the Grace user functions can simply be copied and
modified.  The hardest part is simply that I don't have experience
with matplotlib, so I don't know how to command it to produce high
quality plots.  Maybe I'll post something on the relax-devel mailing
list about this.

Regards,

Edward

On 26 March 2014 21:45, Troels Emtekær Linnet <tlinnet@xxxxxxxxxxxxx> wrote:
> Hi Edward.
>
> I would advise not to save Rex as a value, since it is correlated with the
> effective field.
>
> Ah ja, that was the problem!
> Theta and the effective field are correlated (obvious).
>
> So one can only interpolate, by determining to ramp either spin lock field
> or offset. I wonder how that looks as a 3D graph?
>
> Best
> Troels
>
> Den 26/03/2014 14.42 skrev "Edward d'Auvergne" <edward@xxxxxxxxxxxxx>:
>
>> Hi,
>>
>> Equation 20 in that paper (http://dx.doi.org/10.1021/cr0404287) is
>> actually the sum of the spin-spin relaxation and Rex!  This is exactly
>> what Troels has been working with recently.  See
>> http://wiki.nmr-relax.com/Matplotlib_DPL94_R1rho_R2eff.  In the
>> relaxation dispersion analysis in relax we should really have two
>> parameters automatically generated for the user, R2 and Rex (defined
>> as R2 = R1rho' + Rex).  These could then be used to generate
>> dispersion plots of R2 or Rex verses Omega_eff.  This might take a
>> while though, in the mean time have a look at Troels' script on that
>> wiki page.
>>
>> Regards,
>>
>> Edward
>>
>>
>>
>> On 26 March 2014 14:22, Justin Lecher <j.lecher@xxxxxxxxxxxxx> wrote:
>> > On 26/03/14 13:39, Edward d'Auvergne wrote:
>> >> Hi,
>> >>
>> >> The key here is what is meant by R2.  There are many different
>> >> definitions.  In relax, the parameter name R2 is defined as:
>> >>
>> >> - In the standard relaxation equations, spin-spin relaxation
>> >> component.  I.e. the part influenced by the spectral density function
>> >> J(w).
>> >>
>> >> - Again in the standard relaxation equations, R2 is used for R2*,
>> >> where R2* = R2 + Rex.  R2* is also defined differently if you look at
>> >> the exponential decay curves or if you looking at peak widths (for the
>> >> later there are additional factors broadening the peaks which add to
>> >> the R2* equation).
>> >>
>> >> - For relaxation dispersion, R2 is currently defined as R1rho', i.e.
>> >> just the parts influenced by the spectral density function.  Here
>> >> R1rho' is the on-resonance part of R1rho, excluding exchange.
>> >>
>> >> R2 = R1rho' + Rex could also be defined, which is probably what you
>> >> are after.  The value of R1rho' + Rex has been named many different
>> >> things by the field and there seems to be no consensus.  I have
>> >> discussed this with Troels at
>> >> http://thread.gmane.org/gmane.science.nmr.relax.devel/5119/focus=5207.
>> >>  This is a long thread with many discussions about implementing this
>> >> as an automatically calculated parameter - which in the end would be a
>> >> great feature.
>> >>
>> >> So which R2 value are you after?
>> >>
>> >> Regards,
>> >>
>> >> Edward
>> >
>> > Hi Edward,
>> >
>> > Principally I am looking for pure R2 without Rex contribution.
>> >
>> > Currently we are using some home made scripts which basically are using
>> > equation 20 in Palmer, Massi, 2006 (PMID 16683750).
>> >
>> >
>> > So for the technical side, how do I proceed in relax? Do I need to fix
>> > the decay for R1 and R1rho and then calculate R2 or just fit R1 and use
>> > my raw intensities from the R1rho measurements for the R2 calculation?
>> >
>> > Thanks,
>> > Justin
>> >
>> >
>> > --
>> > Justin Lecher
>> > Institute of Complex Systems
>> > ICS-6 Structural Biochemistry
>> > Research Centre Juelich
>> > 52425 Juelich, Germany
>> > phone: +49 2461 61 2117
>> >
>> >