I am currently refining a dataset of GDH, which is shown to have D3 symmetry.
Using cryosparc gives me good results, however cistem seems to treat the symmetry wrong.
3D reconstructions visually differ from 2D classes and appear to have D6 symmetry applied rather D3.
Even if I take the alignment file from cryosparc and only generate 3D the reconstruction is wrong.
If I set the symmetry to C1 during import, however, everything looks reasonable.
Probably the symmetry definition are different between the two programs. If you take a good reconstruction, and run align_symmettry in the command line it should align it. You can then reimport that and continue.