How to trim SAXS data that it won't come useless

[Mika]

I am using one program that can only process SAXS data containing 256 or less data points. When SAXS data set is trimmed to 256 data points and plotted, it looks very different to what it was at the small angles. I am not sure can I still use the data. Could someone give some advices?

Here are more details:
So, I am using ENSEMBLE (can be found from here http://abragam.med.utoronto.ca/~JFKlab/) which can calculate protein structures. Input data can be chemical shifts, R2 relaxations rates, SAXS scattering curve, etc.

ENSEMBLE contain three external programs and one of them is CRYSOL. CRYSOL is used to calculate SAXS restraints and many ATSAS forum users probably know this program already (it is distributed by EMBL, like ATSAS). So, this program will be using my trimmed SAXS data.

I have an intrinsically disordered protein and it means that it does not have a well-defined 3D-structure. Therefore, it is harder to analyze with SAXS, but it can be done. Also, my curve looks little different to globular proteins. Here is an illustration of my averaged and subtracted SAXS curve containing 1043 data points:

(Check figure "Original SAXS data")

Then I trim the SAXS data and take out every 5th data point. Now SAXS data set has 208 points and the curve looks like this:

(Check figure "Modified SAXS data")

Original and trimmed SAXS data are very similar at big angles, but at the small angles, there is a change. The curve is gone and it resembles a line. I am not sure have I now lose too much information. If I put this trimmed SAXS data in the ENSEMBLE program, I am not sure can CRYSOL calculate SAXS scattering curve which would fit to this trimmed experimental curve.

I hope that my explanation was clear? Can someone give some advices?

[AL]

I am using one program that can only process SAXS data containing 256 or less data points.
...
ENSEMBLE contain three external programs and one of them is CRYSOL.

I guess your question about rebinning data is already answered.
But where does the 256 points limitation come from? CRYSOL can compute up to 5000 points in prediction mode; in fitting mode CRYSOL internally rebins predicted curve to the grid of the experimental data.

Have you tried EOM online?