Difficulty resolving subregion, even with local refinement

Cryo-EM Data Processing
1

Hi,

I am currently working on refining a globular, medium-sized protein-RNA complex (~210 kDa). While refining my particle stack, the GSFSC resolution of my volume improves; however, there is a region that does not seem to improve with further refinements. I am confident in my particle picking, but am concerned whether the cleaning of my particle stack may be the point of failure. I have utilized different workflows – initial 2D classification, direct ab initio reconstruction, and formation of junk volumes – but all methods have led me to a similar conclusion. While density is present, it becomes visible only when the threshold is altered. When creating a mask for local refinement of the low-resolution region, the resultant volume looks identical. Running a 2D classification of the final particle stack, it appears that the region is less resolved. Subsetting the particle stack has not shown any improvement in refinement. Any help would be appreciated!

cryosparc-1
cryosparc-11920×1080 154 KB

2

I would start by increasing your box size, it doesn’t seem large enough for the particle you are dealing with. Possibly around 340 pix should be good, just re-extract them and run another refinement. Is the region you are interested in toward the edge of the box on the bottom of the structure you are showing in the two map views?

3

As @jcoleman suggestions, increase your box size. I’d suggest 400-450. If you’re really after that less clear region at the bottom of your 3D, I’d actually suggest 512 pixel boxes to give a focussed mask and Local Refinement room to work with the adjusted fulcrum.

Also, regarding resolution - you’re losing your (delocalised) high resolution information with a box that tight.

1 Like