Dear all, I’m encountering some shift problems when creating a mask from a pdb file.
What I do is the following:
1) Superimpose pdb on the cryo-EM map
2) Create a mrc map from a pdb structure with mol map in chimera or: e2pdbmr.py --apix X --box XX --res X in.pdb out.mrc
3) Create a mask: relion_mask_create --i out.mrc --o out_mask.mrc --ini_threshold X --extend_inimask --worth_soft_edge X
While the pdb and the corresponding mrc map from step 1) and 2) are perfectly on top of my cryo-EM map, the mask created in 3) is always shifted. And this doesn’t make any sense to me.
After trying all kind of things, I realized the shift observed for the mask after step 3) can be influenced by the apix used in step 2), although it has no influence on the position of the map rendered in step 2). And I can’t find the logic behind this. Perhaps too simply to see...
For the project I’m currently working on the shift is small. The results from post processing in relion suggest that the input refined map was shifted to fit the mask as the final processed.mrc is at the same xyzzy position as the mask but at a different position from the input. And I can tell from the postprocessed_masked.mrc that nothing was “cut” out.
So ist this something I should even bother about since during post-processing the input and mask appear to be superimposed?
Still would like to understand why relion_mask_create induces this xyzzy shift.
Any suggestions are highly welcome.
Best,
Cristina