I am currently processing a dimeric complex and am encountering a problem after symmetry expansion and re-extraction of the particles.
I first refined my complex in C1, using homogenous refinement, followed by an homogenous refinement in C2.
Each monomer have strong conformational heterogeneity, so I proceeded with C2 Symmetry Expansion, and another round of homogenous refinement in C1 with a mask covering only one monomer.
I then shifted my particles using the Volume Align Tools to recenter the particles on only one monomer, and performed another round of homogenous refinement in C1.
Then I reextracted the particles (duplicated), refined them again, and now i am getting weird FSC curves that won’t cross the 0 line …
Like this:
This is expected. When you perform symmetry expansion, you are effectively duplicating your particles but changing the orientation of the copies.
This won’t cause a problem in local refinement (as the orientations of the copies are different, you won’t get spurious correlations), but it will cause a problem in homogeneous or NU refine, where complete orientational search is performed.
I would suggest performing local refinement immediately after symmetry expansion (and you can then perform subparticle extraction if you wish).
Just chiming in; as of CryoSPARC v4.5 released May 7, 2024, refinements with global pose search will raise a warning if symmetry expanded particles are detected in the input particles.
