[RASMB] sedfit vs SOMO vs structure

[Leech, AP]

Hi Borries,

I don't often see RMSD as low as 3e-3 (0.003), though the residuals are
random. I would say 0.005 is a "good day" for our machine.

Andrew

On 08/07/2011 13:47, Borries Demeler wrote:
>>
>> Hello RASMB,Sedfit c(s) analysis of a protein I'm working with shows a 30 =
>> kDa peak with S =3D 2.60 and f/f0 of 1.2, that corresponds to monomer.SOMO =
>> in Ultrascan using the pdb file for this protein predicts a 30 kDa monomer =
>> at 2.42, with f/f0 of 1.19.  It's good agreement with the sedimentation =
>> velocity results.However, the crystal structure, SEC, etc. indicate this =
>> protein should be a dimer.Starting with the default value of 1.2 for the =
>> frictional ratio and floating in sedfit returns a value of 1.2, with MW =
>> that corresponds to monomer.  RMSD =3D 0.009756Doubling this value to 2.4 =
>> and floating gives a value of 2.3, and assigns a mass that corresponds to =
>> dimer to the main peak in the c(s) analysis.  RMSD =3D 0.009324Thus there =
>> appear to be two energy minima for the frictional ratio, depending which =
>> initial value is used.  The RMSD value is lower when the higher value of =
>> f/f0 is used.Would you, on the basis of RMSD values, assign monomer or =
>> dimer to this peak?Thanks,Mark ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~=
>> ~~~~~~~~~~~~~
>> Dr Mark Agacan, Scientific Officer for the Division of Biological =
>> Chemistry and Drug Discovery,
>> Wellcome Trust Biocentre, College of Life Sciences, University of Dundee, =
>> Dundee, DD1 5EH=20
>> Tel: +44 1382 386095    Fax: +44 1382 345764    Mobile: 07525 451 117
>> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~=20
>
> Mark, is this absorbance or interference data? If absorbance data I
> find the RMSD values about 2-3 times too high to be representative. So I
> would question the quality of the data or the fit, or both. On absorbance
> data we typically get about 2e-3 - 3e-3 for RMSD at 280 and 230 nm. Do
> you get random residuals throughout the fit? If so, I would check the
> instrument. If the fits doesn't produce random residuals, I would question
> your fit and model.
>
> Regards, -Borries
> _______________________________________________
> RASMB mailing list
> RASMB at rasmb.bbri.org
> http://rasmb.bbri.org/cgi-bin/mailman/listinfo/rasmb

-- 
Dr Andrew Leech                   *  Laboratory Head
Technology Facility               *  Molecular Interactions Laboratory
Department of Biology (Area 15)   *  Tel   : +44 (0)1904 328723
University of York                *  Fax   : +44 (0)1904 328804
PO Box 373,  York  YO10 5YW       *  Email : apl3 at york.ac.uk
EMAIL DISCLAIMER: http://www.york.ac.uk/docs/disclaimer/email.htm