[RASMB] minimum and maximum

Mon Feb 14 09:14:00 PST 2005

Chin,
What you say sounds reasonable.  Unfortunately, it is really not 
possible to answer this reliably in a general way without knowing the 
data.  That the program crashed could mean many things (e.g. that you 
found a bug, which I appreciate, that your computer is limited in 
memory, or that something is not set up right) and makes me reluctant to 
recommend how to proceed, although it seems right.  If you want you 
could send me screenshots of sedfit showing the data, the distribution, 
and the parameter box.  You can reach me directly at pschuck at helix.nih.gov.
Peter

Qin Zou wrote:

>Hi, Peter,
>Thanks for the input. Now I still have one thing unclear. At smax, the c(s)
>seems never come down to the baseline. There is always an upward trend that
>looks like a half peak. In the SEDFIT tutorial, it suggests that smax should
>be increased to get even large species if there is an increase of
>distribution function. Now as I mentioned before that I set smax at 614s,
>the distribution still increases at smax. Beyond that, the program crashed.
>So if I only want to see the integration of total polymer species, can I
>just integrate all the way to smax without worrying the increase of c(s) at
>smax? Thanks in advance.
>
>Chin
>
>-----Original Message-----
>From: rasmb-admin at server1.bbri.org [mailto:rasmb-admin at server1.bbri.org] On
>Behalf Of Peter Schuck
>Sent: Wednesday, January 26, 2005 10:14 AM
>To: RASMB at server1.bbri.org
>Subject: [RASMB] minimum and maximum
>
>----------------------------------------------------------------------------
>------
>The older archived RASMB emails can be found at:
>http://rasmb-email.bbri.org/rasmb_archives
>and current archives at
>http://rasmb-email.bbri.org/pipermail/rasmb/
>Search All the Archives at:
>http://rasmb-email.bbri.org/rasmb_search.html
>----------------------------------------------------------------------------
>------
>
>Hi Chin and Joris,
>
>my experience with large and small particles is the following:
>
>1) as for the maximum, there is in principle no upper limit in 
>c(s).  However, there is a limitation due to the experimental data that are 
>loaded in sedfit.  For example, if you want to observe a speices with an 
>s-value so large that for the first scan it would already be at the bottom 
>of the cell, or outside your fitting range, then it doesn't make sense and 
>sedfit will prevent you from using such an s-max.  An error message should 
>appear with the maximal s-value that you can possibly observe for the given 
>data.  But if you run slow enough, that's not limiting.  I've looked at 
>size-distributions of bacteria with 10,000  - 100,000 S.  For very large 
>particles, you may want to consider switching to the ls-g*(s), since 
>diffusion can likely be neglected.
>
>2) for the smallest species, we can routinely see the sedimentation of 
>buffer salts (and measure their s-value...), and to characterize things in 
>the couple hundred Dalton range is absolutely feasible.  I would definitely 
>recommend to go to 60,000 rpm and a long (maybe 10 mm) solution 
>column.  The question of equilibrium versus velocity is simple to answer, 
>since they will be essentially the same - except that if you regard it as a 
>velocity you would acquire all the data towards the approach of 
>equilibrium, which is extremely valuable information.  Of course, it will 
>be best to also let the solution actually come to equilibrium, which 
>shouldn't take too long.  Including the approach to equilibrium in the 
>analysis is helpful, in particular, for resolving different species, which 
>may be difficult to unravel from the equilibrium alone.   You can find some 
>examples for the shapes of sedimentation profiles in Biophys J (1998) 
>74:466, but the c(s) can work well in that range, too.  Examples for 
>approach to equilibrium analyses with c(s) and c(M) can be found in 
>Biophys. J. (2001) 81:371
>
>The problem with velocity that Borries referred to is probably the 
>expectation to pull a traditional boundary, clear the meniscus and get 
>plateaus, which is required, for example, for an analysis with the van 
>Holde-Weischet method.  That would not work.  But if you regard a velocity 
>run as simply observing the time-course of sedimentation, it should work 
>perfectly fine if you use Lamm equation solutions in the analysis and you 
>can get perfectly good s and D values.  If by any chance your two molecules 
>have significantly different absorption spectra, I would try to take 
>advantage of that in multi-signal detection.  For interacting systems with 
>known reaction scheme, the Lamm equation solutions incorporating reaction 
>terms should be applicable.
>
>One issue that in my experience makes more difficulties with very small 
>species is the v-bar problem, maybe due to the larger surface-to-volume 
>ratio and relative contributions from solvation.  But if the problem is 
>such that the 500 Da and the 1500 Da can be characterized separately, 
>first, then you could take the buoyant molar mass of the individual species 
>(together with the known mass from mass spec) to calculate an effective 
>v-bar which is correct for the experimental conditions.
>
>Best,
>Peter
>
>
>At 11:22 AM 1/26/2005 +0100, you wrote:
>  
>
>>---------------------------------------------------------------------------
>>    
>>
>-------
>  
>
>>The older archived RASMB emails can be found at:
>>http://rasmb-email.bbri.org/rasmb_archives
>>and current archives at
>>http://rasmb-email.bbri.org/pipermail/rasmb/
>>Search All the Archives at:
>>http://rasmb-email.bbri.org/rasmb_search.html
>>---------------------------------------------------------------------------
>>    
>>
>-------
>  
>
>>Dear all,
>>
>>Some time ago I asked around for the highest measurable molecular weight 
>>by analytical ultracentrifugation. This time a colleague asked me whether 
>>it is possible to measure host-guest complexes by analytical 
>>ultracentrifugation in which the host is around 1500 Da and the guest 
>>around 500 Da. I checked literature on this more supramolecular chemistry 
>>related research (in contrast to the more biochemical research normally 
>>done by AU) and so far without any luck. Could someone help me with a 
>>lead? To me it seems to be possible, although one needs to spin the sample 
>>pretty fast. Apart from this question: what would be the experiment of 
>>choice? Sedimentation velocity or sedimentation equilibrium?
>>Thanks!
>>Kind regards,
>>
>>Joris Beld
>>Swiss Federal Institute of Technology
>>Hilvert Group
>>_______________________________________________
>>RASMB mailing list
>>RASMB at rasmb-email.bbri.org
>>http://rasmb-email.bbri.org/mailman/listinfo/rasmb
>>    
>>
>
>_______________________________________________
>RASMB mailing list
>RASMB at rasmb-email.bbri.org
>http://rasmb-email.bbri.org/mailman/listinfo/rasmb
>
>
>  
>