Hi,
I am now performing distance distribution analysis using primusqt.
I have some troubles:
1. I managed to obtain the plot which seems to have smooth shape and reasonable Rg compared with the one I obtained from Guinier analysis (Rg = 63.99) (Fig. 1). However, If I unchecked rmax = 0, the value of p(r) at rmax was far from 0 (Fig. 2).
2. I then increased the value of Dmax to make the value of p(r) at rmax come closer to 0. I could get the smooth graph, but in this case, the message "Warning: Dmax*smin greater than pi" appeared and moreover, Rg (72.49) did not fit with the Rg obtained from Guinier analysis (63.99) (Fig. 3).
I want to use this pr function for ab initio modeling by DAMMIN.
I've learned from previous posts that it is not good to use the input file with the warning message for ab initio modeling.
Also, I've read that dmax value would largely affect the result of bead modeling, so I was worrying that I might get wrong bead models if I used pr function which I have shown in Fig. 1.
How can I obtain accurate dmax value in this case?
Thank you so much.
How can I decide Dmax correctly?
How can I decide Dmax correctly?
 Attachments

 Figure 2.png (40.58 KiB) Viewed 205 times

 Figure 1.png (40.88 KiB) Viewed 205 times
Re: How can I decide Dmax correctly?
I'm sorry, this is Figure 3.
 Attachments

 Figure 3.png (41.94 KiB) Viewed 199 times
Re: How can I decide Dmax correctly?
Dear Anna,
you have very low intensity scattering (negative Log scale), is this curve coming from SECSAXS?
If yes, try to better select the range for integration.
Your data between 0.018 and 0.02 curve negatively, this will affect both Rg and Dmax. Therefore, unless you have a good structure to guide you, I would not trust too much the envelope from DAMMINDAMMIF.
HTH
best regards
Adriana
you have very low intensity scattering (negative Log scale), is this curve coming from SECSAXS?
If yes, try to better select the range for integration.
Your data between 0.018 and 0.02 curve negatively, this will affect both Rg and Dmax. Therefore, unless you have a good structure to guide you, I would not trust too much the envelope from DAMMINDAMMIF.
HTH
best regards
Adriana
Re: How can I decide Dmax correctly?
Dear Adriana,
Thank you for your advice.
The curves are not from SECSAXS, but are from normal SAXS.
Unfortunately, I do not have a crystal structure of the protein I used for this SAXS measurement, so I think I should try the measurement again.
Is there any way to increase the scattering intensity other than increasing the concentration?
I would appreciate it if you could give me advice again.
Yours sincerely,
Anna
Thank you for your advice.
The curves are not from SECSAXS, but are from normal SAXS.
Unfortunately, I do not have a crystal structure of the protein I used for this SAXS measurement, so I think I should try the measurement again.
Is there any way to increase the scattering intensity other than increasing the concentration?
I would appreciate it if you could give me advice again.
Yours sincerely,
Anna
Validate your MW, try DARA, DATCLASS
Fig. 3 looks like your D_{max} is about 265 A. Indeed your s_{min} = 0.013 A^{1}, so formally you cannot correctly estimate D_{max} > 242 A.
If D_{max} = 265 A fits your expectations (for the MW computed from the sequence) make sure 'Rmax=0' is checked, save and give DAMMIN/DAMMIF a try.
Before starting any modelling make sure the MW estimated from the SAXS data fits the expected MW.
If you don't know how to estimate D_{max} you may submit your data to DARA and check what the D_{max} values of proteins with similar scattering patterns are. You may also use the command line tool DATCLASS to estimate MW and D_{max}.
If D_{max} = 265 A fits your expectations (for the MW computed from the sequence) make sure 'Rmax=0' is checked, save and give DAMMIN/DAMMIF a try.
Before starting any modelling make sure the MW estimated from the SAXS data fits the expected MW.
If you don't know how to estimate D_{max} you may submit your data to DARA and check what the D_{max} values of proteins with similar scattering patterns are. You may also use the command line tool DATCLASS to estimate MW and D_{max}.