Process Ctf refined/polished particles from Relion

Dear All,

I am trying to process a dataset by doing the earlier steps in RELION and 3D classification/refinement in cryosparc. If I directly import particles after extraction/selection from RELION, the refinement in cryopsarc went on well and returned good results. However, after I did CtfRefinement and Bayesian polishing in RELION and import the polished particles, the refined structure, although report higher resolution, is with wired pattern, as shown in the attached figures. I tried refining both sets in relion and no issue. I tried doing the Ctf and polishing in RELION 3.0 or 3.1 and get similar wired pattern. The ctfrefinement and polishing are done simply following the RELION tutorial. I am writing to ask if someone may have encounter similar issues and know how to solve it. Thank you a lot. Highly appreciate any suggestions and comments.

Yang

Hi Yang,

After polishing in Relion, detector defects can be amplified in a small subset of the particle images. Your problem seems similar to what I see when that occurs. These are easily removed using another round of 2D classification after the polishing step.

If that is not the problem, it could be that CTF refinement failed, which can happen if the reference is not very high resolution.