Reliability of Crysol theoretical curves

Calculation of SAXS and SANS profiles (CRYSOL, CRYSON), superposition of models (SUPCOMB, DAMAVER, DAMCLUST), database DARA
Post Reply
Message
Author
ClaFusis
Member
Posts: 3
Joined: 2016.12.15 13:10

Reliability of Crysol theoretical curves

#1 Post by ClaFusis » 2016.12.15 14:44

Hello everybody,
my group and me took several SAXS data of our protein (Archaeglobus Fulgidus Ferritin, PDB 1S3Q) under assembled and disassembled conditions. (This protein has a hollow sphere structure and is made of 24 subunits that dissociate into dimers).

In this way we had different curves of what we believe to be dimers or 24mers or a mixture of them with unknown proportions.
My task is to compare these experimental data with the theoretical scattering curves calculated with CRYSOL and try to find these proportions with a Matlab script I'm writing (in details, I write my data as a linear combination of the theoretical patterns and I try to find these coefficients).

My question is: how much are the theoretical curves reliable for this kind of analysis? The linear combination, in fact, strongly depends on the theoretical curves.
In particular, I have strong doubts on the 24mer theoretical pattern that has very deep minima: our protein pattern agrees with it but doesn't have that sharp shape. You can easily see this in this figure, in which I have plotted with Matlab the theoretical patterns of the 24mer (in black, dashed lines) and the dimer (in black, continue lines) and my experimental data of a 100% 24mer protein (blue dots) and a 100% dimer (red dots).
SAXScurves.png
SAXScurves.png (61.08 KiB) Viewed 4534 times
As you can see, the dimer experimental curve (scaled with a costant to fit the theoretical) agrees very well with its theoretical pattern. Instead, the 24mer has more or less the same shape but has smoother peaks.
This is a huge problem for my kind of analysis because, with these theoretical curves, my program will tell me that the experimental dimer is a 100% dimer (as it should be), instead the 24mer will be something like 60% 24mer and 40% dimer (but I know it's not!). And in more complicated samples it will just not be realiable!

What would you suggest me to do?
Thanks a lot for your attention! :D :D :D

User avatar
AL
Administrator
Posts: 686
Joined: 2007.08.03 18:55
Location: EMBL Hamburg, Germany
Contact:

Re: Reliability of Crysol theoretical curves

#2 Post by AL » 2016.12.20 15:04

ClaFusis wrote:My task is to compare these experimental data with the theoretical scattering curves calculated with CRYSOL and try to find these proportions with a Matlab script I'm writing (in details, I write my data as a linear combination of the theoretical patterns and I try to find these coefficients).
Have you tried OLIGOMER?

ClaFusis
Member
Posts: 3
Joined: 2016.12.15 13:10

Re: Reliability of Crysol theoretical curves

#3 Post by ClaFusis » 2016.12.20 15:38

AL wrote:
ClaFusis wrote:My task is to compare these experimental data with the theoretical scattering curves calculated with CRYSOL and try to find these proportions with a Matlab script I'm writing (in details, I write my data as a linear combination of the theoretical patterns and I try to find these coefficients).
Have you tried OLIGOMER?
No, I didn't, since I would like to do this analysis by myself with Matlab (I always analyze data in this way)!

SaxsMax
Administrator
Posts: 349
Joined: 2007.08.03 19:16

Re: Reliability of Crysol theoretical curves

#4 Post by SaxsMax » 2016.12.20 16:27

The smoothing of the experimental pattern could be either due to polydispersity (which you beleive is not the case)
or due to some flexibility in the positions of the monomers within the 24mer.

Possibly, you could use the experimental profiles of the 100% 24mer and of the 100% dimer to deconvolute the other curves.

ClaFusis
Member
Posts: 3
Joined: 2016.12.15 13:10

Re: Reliability of Crysol theoretical curves

#5 Post by ClaFusis » 2016.12.21 11:46

SaxsMax wrote:The smoothing of the experimental pattern could be either due to polydispersity (which you beleive is not the case)
or due to some flexibility in the positions of the monomers within the 24mer.

Possibly, you could use the experimental profiles of the 100% 24mer and of the 100% dimer to deconvolute the other curves.
I'm really sure that our sample is something like 100% 24mer! If this smoothing is due to the flexibility in the positions of the monomers, what do you think I should do?
Anyway, I'll try to do what you said (even though I would prefer to have a reliable theoretical 24mer) and I'll let you know! Thanks!! :D

Alex
Active member
Posts: 692
Joined: 2007.08.09 21:10
Location: Planet Earth

Re: Reliability of Crysol theoretical curves

#6 Post by Alex » 2017.01.10 13:05

Perhaps, analysis of SAXS data as done in the following article can help you:
http://www.pnas.org/content/108/16/6409.long

Alex

Post Reply