I know symmetry expansion has been addressed in several other posts, but I still cannot find the answer to my specific question.
I have a homohexameric protein with D3 symmetry. I refined with D3, performed symmetry expansion, then performed 3D classification with a mask around the asymmetric unit (with force hard class on). I have identified asymmetric units that are disordered and I want to remove all particles that have 1 or more disordered asymmetric units from the original particle stack. I do not see a way to do this without scripting, but if I am missing something please let me know.
I have selected the classes that have ordered asymmetric units then performed local refinement using a mask around the full particle and just the asymmetric unit, and this works well as it should, and it may be enough/sufficient, but it would still be nice to have an unexpanded particle stack that contains only particles where all 6 asymmetric units are well-ordered
See this thread (and the script at the end): Split output for symmetry expansion - #3 by olibclarke
I think this CS tools script should to what you are after (we have used it for basically the same purpose)
Hey Oli,
This is indeed what I need. It would also have been doable by converting the particles_all_classes.cs file to a star file if the star file retained the UID field, but I tested and it did not. Really all you need to do if that worked was to identify the classes with the ordered asymmetric unit, or in your case, ligand bound, filter the star file for just those classes (if multiple), then filter the star file for UIDs that appear N times (6 in my case). The star file could then be re-imported as a particle stack and reassigned to the original micrographs. Alas, this didnt work.
I’ve never used cryosparc tools, so would it be okay if I reached out to you via email to get some help with implementing this workflow?
also works to just refine without symmetry and classify globally. For instance, in my case, I have a map where one of the subunits is missing/highly disordered, so I used that as the ref for C1 refinement, which did a good job of aligning most particles with their disordered subunit in the same position, then just did 3D class after and was able to clean up the dataset that way, then did D3 refinement with the truly symmetric particles. Symmetry expansion is nice though and it would be even better if this workflow was less cumbersome.
Sure - I would say have a go with some of the basic scripts & tutorials available via the guide, try to get them to work, and then if you still can’t get this one to work feel free to reach out either via email or DM and I’ll try to help if I can!
A pretty simple way to do it, without scripts, is to iterate remove duplicates jobs, to get to specific to parent particle sets you are after. Just be careful your final stack doesn’t contain any duplicates (I usually do a sanity check with additional remove duplicate job)
It would be nice if cryoSPARC had a more comprehensive revert symmetry expansion tool that would allow one to selectively retain priors/metadata from desired symm expanded particle subsets after focused 3D classification - doing this to reconstruct different subunit assemblies/conformer combos in symmetric particles would be handy. I came up with rudimentary way to do this in the past, since I consider myself more or less “coding illiterate” - Antiviral drug recognition and elevator-type transport motions of CNT3 | Nature Chemical Biology