Re: [task #7858] Make it possible to submit CPMG experiments for BMRB -- October 06, 2014 - 16:50

Hi Eldon,

For your own information, as this may not be considered for a while,
or even ever, Troels has added a lot of data published by Kaare into
the relax sources at
http://svn.gna.org/viewcvs/relax/trunk/test_suite/shared_data/dispersion/
in the directories including *KTeilum*.  This is generally the input
relaxation data though you will find a few directories with saved
relax results files, for example in KTeilum_FMPoulsen_MAkke_2006/.
Other resources which may be of interest include the different
dispersion models and all their parameters:

http://www.nmr-relax.com/manual/Dispersion_model_summary.html

The analysis protocols used, for example the automated analysis in relax:

http://www.nmr-relax.com/manual/relaxation_dispersion_auto_analysis.html

The dispersion curve insignificance concept which would somehow have
to be included in the deposition:

http://www.nmr-relax.com/manual/Dispersion_curve_insignificance.html

The essential concept of clustering of spin systems:

http://www.nmr-relax.com/manual/clustered_relaxation_dispersion_analysis.html

The different software for dispersion:

http://www.nmr-relax.com/manual/Comparison_dispersion_analysis_software.html

There is also a few additional details required, for example the
correction for the R1rho data so that the correct ground state
chemical shift and offset are used - something which is not yet
implemented in relax.  Some more are listed in the relax to do list at
http://www.nmr-relax.com/manual/do_dispersion_features_yet_be_implemented.html.

These are all from the relax manual, the PDF version of which is of
much higher quality (http://download.gna.org/relax/manual/relax.pdf).
Also, if you have a look at
http://www.nmr-relax.com/api/3.3/target_functions.relax_disp.Dispersion-class.html#__init__,
you will see that the relaxation data used for dispersion is
characterised by the following properties:

    Ei: The index for each experiment type.
    Si: The index for each spin of the spin cluster.
    Mi: The index for each magnetic field strength.
    Oi: The index for each spin-lock offset. In the case of CPMG-type
data, this index is currently always zero.
    Di: The index for each dispersion point (either the spin-lock
field strength or the nu_CPMG frequency).
    Ti: The index for each time point. This is currently unused but
might change in the future.

So you could see that multiple saveframes will be required for each
relaxation data set which is characterised by at least the experiment
type, magnetic field strength, spin-look or hard pulse offset, and
dispersion point (spin-lock field or CPMG frequency).  This part
should be relatively easy to implement with the current NMR-STAR 3.1
format (http://www.bmrb.wisc.edu/dictionary/).  But the dispersion
models and parameters will probably require a new saveframe to be
designed and implemented in both the BMRB and in relax/bmrblib.

The ADIT-NMR backend at http://deposit.bmrb.wisc.edu/bmrb-adit/ would
probably also need some changes, as I know that many of the details in
the current model-free NMR-STAR formatted files created by relax are
not accepted by the system any more.  For example the NMR-STAR
formatted test file at
http://svn.gna.org/viewcvs/relax/trunk/test_suite/shared_data/model_free/OMP/bmrb.star?view=log
or 
http://svn.gna.org/viewcvs/*checkout*/relax/trunk/test_suite/shared_data/model_free/OMP/bmrb.star.
It seems as if the system does not accept a fully formatted NMR-STAR
file as valid input any more.

Regards,

Edward


On 6 October 2014 15:10, Eldon Ulrich <elu@xxxxxxxxxxxxx> wrote:
> Hi,
>
> At BMRB, we would be very interested in working on modeling and archiving
> the relaxation dispersion data. We can archive the data without the modeling
> initially. We would really like to have the data deposited. If anyone would
> be willing to send us examples of the data, we can begin the modeling
> process.
>
> Best,
> Eldon
>
>
> On 10/6/14, 7:08 AM, Edward d'Auvergne wrote:
> > Hi Troels,
> >
> > The support for handling relaxation dispersion data for BMRB
> > deposition has been a long running issue.  I have included Eldon
> > Ulrich from the BMRB in the CC list as he would be the person to
> > discuss this with (note this is the relax-devel public mailing list
> > which is permanently archived).  Since discussions at the ENC
> > conference in 2009, Eldon and I have jointly developed both the
> > support in relax and the required NMR-STAR saveframes and tags for
> > fully and properly supporting the deposition and reading of model-free
> > data.  This includes the base relaxation data and its metadata, all
> > parameters used such as CSA and dipolar coupling information, the
> > diffusion tensor, and the internal model-free motions (as well as
> > citations, the molecular assembly, experimental details, etc).  See
> >
> > http://www.nmr-relax.com/api/3.3/specific_analyses.model_free.api.Model_free-class.html#bmrb_write
> > and
> > http://www.nmr-relax.com/api/3.3/specific_analyses.model_free.api-pysrc.html#Model_free.bmrb_write.
> >
> > As for relaxation dispersion data, I think I have discussed this with
> > Eldon in person (and possibly also with Art Palmer).  But the
> > relaxation dispersion analysis in relax was at the time not in a
> > useful state, unlike today where relax is now probably faster and more
> > comprehensive than all other dispersion software out there.  I do not
> > know the current state of the BMRB database with respect to relaxation
> > dispersion data, but Eldon should be able to help.  If the support is
> > not currently there, then we would need to follow the same development
> > path as with the model-free analysis.  This would involve joint
> > discussions with Eldon about what NMR-STAR saveframes and tags need to
> > be developed to support the data.  These would then have to be added
> > to the BMRB.  And we would need to extend both relax and the bmrblib
> > library (https://gna.org/projects/bmrblib/), which used to be part of
> > relax and which uses Jurgen Doreleijers' Pystarlib library
> > (https://code.google.com/p/pystarlib/).  The STAR format is relatively
> > simple (http://www.bmrb.wisc.edu/formats.html) but Jurgen's library
> > handles that part for you.  Troels, your coding skill are now so
> > advanced that you should have no problem adding this support to the
> > relaxation dispersion analysis in relax and to also extend the brmblib
> > which is really also a part of relax.  It would only be a couple of
> > weeks to a month to develop.  The hardest part may be to discuss and
> > have all the required tags added to the BMRB itself.
> >
> > However for you Troels, I would recommend that you consider this for a
> > later project, as I know you have some other important deadlines to
> > meet first.  I believe there is strong interest on the side of the
> > BMRB to have this type of data deposited, and with relax we have the
> > best starting point for this.  Kaare Teilum has a lot of high quality
> > CPMG and R1rho data which could form the basis of the initial BMRB
> > depositions.  From Kaare's point of view, I don't know how beneficial
> > this would be.  Anyway, you should discuss this with your supervisor
> > Kaare and determine if this short project is worth your time, after
> > your other deadlines have been met.
> >
> > Regards,
> >
> > Edward
> >
> >
> > On 3 October 2014 18:31, Troels E. Linnet
> > <NO-REPLY.INVALID-ADDRESS@xxxxxxx> wrote:
> > > Update of task #7858 (project relax):
> > >
> > >                    Status:             In Progress => Gone walkabout
> > >
> > >      _______________________________________________________
> > >
> > > Follow-up Comment #6:
> > >
> > > This feature will not be implemented, as it requires a large re-write of
> > > data
> > > structures.
> > >
> > > To implement the funvtion, it would require a re-write of the relax_data
> > > bmrb_write(star) function, and proper handling of cdp.ri_ids.
> > > It was also not readily possible to find examples of submitted CPMG data
> > > in
> > > the BMRB database.
> > > This makes it hard to develop, and even ensure that BMRB would accept the
> > > format.
> > >
> > >
> > >      _______________________________________________________
> > >
> > > Reply to this item at:
> > >
> > >    <http://gna.org/task/?7858>
> > >
> > > _______________________________________________
> > >    Message sent via/by Gna!
> > >    http://gna.org/
> > >
> > >
> > > _______________________________________________
> > > relax (http://www.nmr-relax.com)
> > >
> > > This is the relax-devel mailing list
> > > relax-devel@xxxxxxx
> > >
> > > To unsubscribe from this list, get a password
> > > reminder, or change your subscription options,
> > > visit the list information page at
> > > https://mail.gna.org/listinfo/relax-devel