Dear Cryosparc users and developers,
I am processing a disc shape protein. I am confused about the results of a 3D classification job :
if the third panel with the x-y plane is correct, the long dimension of the protein should align with the x axis and the y axis instead of the z axis as shown in the first two panels…
Am I missing something?
Thank you for helping me out!
The z refers to the side - the edge of the box - not the dotted line. So the first image is the y-z plane, the second is the x-z plane, and the third is the x-y plane (with z projecting out of page), all sliced through the center - does that make it clearer?
Thank you very much. So, the origin of the coordinate system is at the corner?
My real problem is that I am trying to align the pseudo D6 symmetry axes for symmetry relaxation and I can’t get it right. Cryosparc didn’t seem to find the correct axes. I tried aligning the molecule in Chimera such that the z axis aligns with the 6 fold and tried both putting the other axis along x and y (in the Chimera coordinate system). I then tried to do refinement with the aligned model with ‘do symmetry alignment’ off but the resulting reconstruction had a different orientation.
Could you suggest some things that I can check along the way to make sure that I am doing it right?
Thanks a lot!
Symmetry alignment can be a bit tricky - sometimes the automatic procedure in Cryosparc gets it wrong. When you say you tried to align it in Chimera what exactly did you do? I would make two copies of the map, rotate one with the appropriate transformation, then resample it on the grid of the original map - is that what you did? You also should be able to apply the appropriate transformation using volume alignment tools
I’m really not sure I see what the problem is - the example image provided looks aligned pretty well to D6 symmetry? It’s not suffering from any of the usual issues with symmetry alignment (mis-centering on the Z axis is the primary one)…
Your image shows views of the centre slice of the box down the X axis, Y axis and Z axis (left, middle and right), so the labels are for the axes not directly facing out of the screen.
Another way to align (rather than Chimera(X) which I always find a bit messy) would be with relion_align_symmetry and re-importing the resulting map when you have an alignment you’re happy with (then turning off symmetry alignment during refinement).
Thank you very much for your replies! Yes, the C6 axis is always well aligned. It’s the other axis that is the problem.
I did do the volume resampling and imported back to Cryosparc. I actually also tried using the Volume Alignment tool in Cryosparc and that didn’t help either even though the resulting volume for both methods looked ok. But again, since I am all confused about where the axes actually are, maybe I’ve got it all wrong!!
How can I make sure that I have aligned the volume to the D6 symmetry axes correctly for Cryosparc procedures?
Hmmm - if the C6 axis is aligned with Z, I’m not sure I see how the orthogonal twofold can be off, or am I missing something? How does your volume look after C1 ab initio & refinement, for reference?
The pseudo symmetry axes are clear in the C1 structure but the resolution is poor.
