[RASMB] SE interference data

[Borries Demeler]

> 
> Questions - 1. Is there any particular reason why higher speed interference data
> can't be fitted? Might the high protein concentration have anything to do with
> it - non-ideality or the steep gradient at high concentrations?
> 2. I am not very familiar with interference data. Is there any special treatment
> for this type of data before analysis?
> 
> I'd appreciate any suggestions and advice during this dark hour of mine
> (although I understand that it is summer and vacation time).

Mitra, 

the issue you are running into is that the gradient is too steep
at the top of your scans and you wont get an exponential response. 

When the gradient is getting too steep, you are going
to have refractive artifacts, and effects such as Wiener
skewing. Check the RASMB archives for articles on Wiener skewing
(http://rasmb-email.bbri.org/cgi-bin/htsearch).

Also, if the IF optics are out of focus you will have problems.
They need to be focussed on the 2/3 plane to minimize such issues.

You can go to higher fringes if the speed was low enough as to 
not create a steep gradient. Of course, with a shallower gradient
you also get less information.

My recommendation would be to fit the data to the fixed MW model in
UltraScan and trim the top of the scans until you get a random fit on
all scans. Then save this setup and apply the more constrained models
like the self-association models.

Good luck, -Borries