Is is possible resolution revolution in SAXS?

[shpark]

Dear Saxiers,

I appreciate your efforts in the structural biology field.
I would like to ask you a question about the resolution problem.
There were a few papers published to high resolution structures using SAXS data, however, I couldn't have heard general methods to solve the structure using SAXS data at atomic resolutions.
Beam intensity might not be a major problem, because as far as I recall, there were trials using XFELs (correlation coincidence maybe similar to SAXS?).
I was just wondering what was causing the resolution limit in SAXS.

Best regards,
Si Hoon

[AL]

Excellent question.
Please keep in mind that SAXS is not a technique that should be primarily used to solve structures of biological macromolecules but rather to study their behaviour (e.g. conformational changes, oligomerization, complex formation) in solution under various conditions (temperature/pH/buffer composition etc.)
But here is a list of reasons why the resolution revolution has not happened yet (in my opinion).

• Improving the signal-to-noise ratio is tricky because of radiation damage.
• Beam position and intensity jitter may contribute to the noise. Finite beam size contributes to smearing (nowadays becoming negligible).
• Ambiguity of SAXS data (enantiomers look identical; "T vs. C" problem) should be taken into account and can be often resolved using additional modelling restrains.
• Proteins and RNA are actually flexible in solution, some are more flexible than others. This can be taken into account by fitting the SAXS data not with a single model but with an ensemble of models.
• The sample is never perfect - there might be a certain amount of unspecific aggregates and/or polydispersity in terms of oligomers. This sometimes can be resolved by using SEC-SAXS but that may reduce the signal-to-noise ratio.
• "Structure factor" (inter-particle interactions) affects data from higher concentrations, there are no good models yet that would allow us to take it into account when modelling.
• It is still poorly understood how to take the hydration layer into account at higher angles. There are various programs that use one or another approximation when calculating the SAXS data from an atomic model: CRYSOL, FoXS, WAXSiS, AQUASAXS, AXES etc. They all produce slightly different WAXS data.

Anything else I forgot?