Hi Soumya,
The issue you are seeing is often due to diffusion tensor problems.
In most such cases, the simple assumptions that the diffusion tensor
is spherical, spheroidal, or ellipsoidal does not hold. Are you
working with a single domain globular protein? Could you be
experiencing partial dimerisation?
Without some of the graphs or more information, it is a little hard to
understand what could be happening. Ideally plots of the R1, R2, and
NOE data at both fields, together with an S2 plot would be needed.
And the final diffusion tensor values. If you would like to
communicate this information, you could create a support request using
the link https://gna.org/support/?func=additem&group=relax and attach
the plots there. But keep in mind that this information will be
forever public, so maybe you would like to abstract away some of the
information (figures in PNG form, no numbers directly given, etc).
Note that for the publication, you really should deposit the final
results to the BMRB to make the information publicly available anyway
(click on "File->Export for BMRB deposition" in the GUI for easy BMRB
file creation).
Another issue could be data inconsistency. Both fine temperature
control and calibration on a per experiment basis are essential for
obtaining reliable relaxation data. This is described in detail in
the relax manual in the relaxation curve-fitting chapter (for example
at http://www.nmr-relax.com/manual/Temperature_control_calibration.html).
You can use Sebastien Morin's consistency testing analysis in relax
to see if you have any data inconsistencies. See the relax manual for
details (the PDF is of much higher quality
http://download.gna.org/relax/manual/relax.pdf, but the HTML version
is at http://www.nmr-relax.com/manual/Consistency_testing.html).
Regards,
Edward
On 25 February 2014 09:03, Troels Emtekær Linnet <tlinnet@xxxxxxxxxxxxx>
wrote:
Dear Soumya.
Would you be able to provide the script you have used for the analysis?
Or did you use the GUI?
That could help find possible "errors".
I guess that it would be something similar to:
http://www.nmr-relax.com/manual/Single_model_free_model_script_mode_sample_script.html
If you have saved the output from relax prompt, I have tried to make a
grep script that convert the
output to a relax script.
http://wiki.nmr-relax.com/Grep_log_file
Best
Troels
2014-02-25 7:25 GMT+01:00 Soumya Joseph <soumya.joseph@xxxxxxxxxxxxx>:
Hi,
I've conducted model-free analysis using the default protocol set up in the
GUI on a protein for which I have data recorded at 600 and 800 MHz. The
order parameters in s2.txt look very strange: regions I know are dynamic
have larger S2 values than those in the "structured" regions. My
understanding is that if the value is close to zero then it is highly
dynamic and if close to 1 => very rigid. My results are the inverse of what
I'm expecting. The calculation ran without any errors and I've checked the
input data for any obvious errors. Has anyone come across this before?
Another strange thing is that the regions I know to be structured (which
seem to have small S2 values) have very large associated errors. The N- and
C-termini of my protein however have higher S2 values but have smaller
associated errors.
Does any one know what could have happened?
Cheers,
Soumya
_______________________________________________
relax (http://www.nmr-relax.com)
This is the relax-users mailing list
relax-users@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-users
_______________________________________________
relax (http://www.nmr-relax.com)
This is the relax-users mailing list
relax-users@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-users
Re: unexpected mf S2 values