Open rshopa opened 4 years ago
Hi,
calculate_Comption
doesn't mean radiation damage, it only means to consider Compton scattering in the diffraction simulation.
For radiation damage, one has to run XMDYN simulation which was mentioned in the paper. The XMDYN executable can only be obtained by contacting the developers to get the licensing to institutes.
We are also exploring the possibility to include other possible open-source photon-mater-interaction codes in this framework. If you have any possible candidates in your mind and would to share it with us, please let us known.
@ejcjason, thank you for the explanation.
I share the idea of extending PMI beyond demo version of XMDYN, shall look for other modules.
Despite demo components of Xraypac included in SimEx, I initially assumed that the images presented in 2017 SPIE Proceedings by @CFGrote were obtained using solely the options from SingFELPhotonDiffractor. Compton scattering is considered radiation damage, too: it ejects electrons hence ionising atoms.
Greetings,
As far as I understood, radiation damage for SPI could be adjusted in parameters of
XMDYNDemoPhotonMatterInteractor()
andSingFELPhotonDiffractorParameters()
, namely random_rotation, uniform_rotation, calculate_Compton.I checked these options and found out that random_rotation deteriorates the image quite significantly. But I still cannot reproduce the difference (without and with radiation damage) as shown in earlier work (C Fortmann–Grote et al., Proceedings of SPIE (2017), in Figure 5), for the sample is the 2NIP protein.
Here's how it looks like in notebook for two cases:
The resulting images (for
poissonize=False
) are much more alike than shown in the article:So, what did I miss? Maybe multiple diffraction patterns? Another parameters set in properties?
I also tried to comment/remove lines (e.g.
calculate_Compton=False
) and tested this on other proteins (5UDC, 3F70, 2I0J etc) and did not observe the difference.