What are the first SAXS data processing steps?

Frequently Asked Questions about ATSAS small-angle scattering analysis program package
Post Reply
Message
Author
User avatar
AL
Administrator
Posts: 686
Joined: 2007.08.03 18:55
Location: EMBL Hamburg, Germany
Contact:

What are the first SAXS data processing steps?

#1 Post by AL » 2010.08.19 12:52

Primary data reduction and analysis steps
for SAXS from biological macromolecules in solution
using the ATSAS package

1) Radial averaging of a 2D image into 1D scattering curve I(s) (or I(q)).

2) Normalization against beam intensity, exposure time, concentration.
These two steps are normally done directly at the beamline.

3) Radiation damage check.

4) Buffer subtraction: subtract the pure solvent signal (buffer) from the solution signal (sample).
Usually the buffer is measured twice: once before the sample and once after. The proper buffer data should be chosen for subtraction: average of both buffers (if they are identical) or only one of them (whichever fits best).
A properly subtracted curve should not have extensive negative parts and should tend to zero at higher angles.

5) Evaluation of the overall parameters (invariants): radius of gyration (Rg), molecular mass, Porod volume.
5.1) Radius of gyration is the average of square center-of-mass distances in the molecule weighted by the scattering length density. It is a measure for the overall size of a macromolecule. Rg and zero angle intensity I(0) can be obtained from the AUTORG program.
5.2) If normalization against concentration was done properly, the molecular mass can be estimated from I(0):
MMsample = I(0)sample * MMref / I(0)ref
where I(0)ref is zero angle intensity of a reference protein with known molecular mass (e.g. BSA).
5.3) The Porod volume can be estimated using the program DATPOROD. Based on the volume one should also estimate the molecular mass which ideally should be close to the value of molecular mass from I(0).
For globular proteins, the hydrated volume in nm3 is about 1.5-2 times the molecular mass in kDa.

6) Data quality. We assume that the solution is monodisperse and homogeneous. If it is known that the solution is polydisperse (e.g. a mixture of monomers and dimers) volume fractions of each component can be determined using OLIGOMER.
AUTORG checks for aggregation (looks at abrupt upwards trend in the lowest angels) and gives a quality estimation based on the linearity of the Guinier region. Please note that depending on (a) the geometry of the beamline and (b) number of experimental points the Guinier region might be too short for estimating Rg of bigger particles.

7) Flexibility check: Kratky plot. If the sample is flexible further steps are not applicable. Use the EOM approach instead (programs RanCh and Gajoe).

8) Merging high and low concentration data; alternatively extrapolation to zero concentration using ALMERGE.

9) CRYSOL is used to compute the scattering pattern from a known high-resolution structure (e.g. from crystallography) and to compare it with the experimental data.

10) The distance distribution function p(r) - compute it manually using GNOM or automatically using DATGNOM. Dmax is an important measure of particle size. Make sure smin < π/Dmax. The Rg computed from p(r) should be close to Rg estimated by the Guinier approximation. Compare your p(r) to the theoretical distance distribution functions and to the p(r) functions of other users (here is another example).

After these steps one can proceed to ab initio modeling and rigid body modeling.

Post Reply