Anisotropy issues: random ring features

Data Processing Strategy
1

Picture
Picture3726×600 194 KB

Hello everyone,

I’m working with a dataset collected on the Krios microscope for a membrane protein, and I’ve encountered an issue that I don’t fully understand.

In both my micrographs and 2D classifications, I’m seeing random rings appear with varying frequencies (as shown in the attached image).
I’ve tried several methods to remove particles affected by these rings, including multiple rounds of 2D classifications and heterogeneous refinement, etc. However, these strategies still result in a map with very strong anisotropy. It seems that some of the rings, especially those with very low frequencies, might be contributing to this problem.

Has anyone experienced a similar issue or have any insight into the possible causes of these rings? I’d also greatly appreciate any suggestions on how to reduce the anisotropy in my final maps.

Thank you in advance for your help.

2

Possibly detector defects - double check the gain reference? if you map the particles from these classes back to the micrographs, do you see anything unusual?

3

Screenshot
Screenshot900×903 202 KB

Thank you for your response @olibclarke .
When mapping the particles from these classes back to the micrographs, they are localized at the two edges of the micrographs (see the screenshot below, where localization occurs at only one edge).
Could this issue be related to the gain reference?

4

It’s certainly possible, although there are no visible issues here. You can try making a gain by averaging some movies (several hundred) using relion_estimate_gain and see if that helps?