[RASMB] recommended max rmsd for achieved equilibrium

[Cole, James]

Dear Frank-
You should look at the residuals plots as a guide to whether you have achieved equilibrium instead of a target value of RMSD. At equilibrium, the the RMSD should correspond to the intrinsic noise level of the optical system. We have implemented a "Match" algorithm in our sedimentation equilibrium analysis program HeteroAnalysis that provides a systematical way to intercompare successive scans while accounting for potential vertical offsets and rotor precession (horizontal offsets).  Neither of these offsets should be an required with XLA absorbance data, but it still useful to overplot all of the residual plots and to look at plots of RMSD vs. time to monitor the approach to equilibrium.  Keep in mind that some systems never come to equilibrium due to slow precipitation at the high concentrations near the base of the cell. You can estimate how much time it should take to achieve equilibrium based on the molar mass, the solvent viscosity and the column height (Van Holde and Baldwin 1958 J Phys Chem 62, 734). I believe that Walt Stafford has an online calculator and it is also available in Sedfit.   However, at this only a lower limit and it can take longer for highly asymmetric molecules and for systems in slow chemical exchange. The equilibrium time goes as (column height)^2. Unless you are looking at lower molecular solutes you should not need more than a 3 mm column to obtain precise molecular molecular weights, so you may have loaded more than necessary in your 2 sector cells, resulting in slow equilibration.  Finally, you can avoid the issue entirely  by doing velocity sedimentation. If you just need to confirm the oligomeric state of the protein this is an easier experiment. 
Regards,
Jim Cole


On Oct 11, 2013, at 10:23 AM, Frank Niesen <dr.frank.niesen at gmail.com>
 wrote:

> Hello everybody,
> 
> I was hoping for advice on what rmsd between successive scans at one speed I should be aiming for to determine if equilibrium is achieved? Below 0.01? Below 0.005? Or dependent on how the residuals are distributed?
> To give an example, I am spinning samples of a protein that forms trimers in 2-sector cells (190 ul/cell). The highest concentrated sample is 200 uM and I waited for about 90 hours, seeing gradual improvement, to currently 0.019 rmsd (I use Sedfit to match, thanks Peter & Co. for the excellent tool!). The differences toward the meniscus and the bottom are about -0.03 and 0.04, respectively. So, the distribution is still "developing"...
> But, how good is good enough, having in mind the demand for answers by my collaborators and my desire to proceed to the next speed for the weekend? ;o)
> 
> Many thanks for any help you can give, and a great weekend for all!
> 
> Frank
> _______________________________________________
> RASMB mailing list
> RASMB at list.rasmb.org
> http://list.rasmb.org/listinfo.cgi/rasmb-rasmb.org