"False" resolution of local refinement

Cryo-EM Data Processing
1

Dear cryosparc community,
I have a cryo-EM map of a flexible complex with reported overall resolution of 2.9A. And I ran local refinement on a rigid part (around 100kDa) trying to improve local resolution. The first iteration shows resolution of 3.5A and shows some featuer of RNA and proteins. However, the second iteration shows resolution of 1.8A but the map looks like 10A. The masked region looks like blob and the region outside the mask is filled with noise.

image
image976×1044 102 KB

Any suggestions what to do to improve/correct the result? Thank you so much.
—Da

2

That is odd and seems like some issue with the mask provided for local refinement.

Did you make the mask using the volume from a previous job with particles + half-maps, and also resample the mask on that ? Another issue could be in the Volumes Tools job with Dilation (number of pixels), soft padding (number of pixels), and Threshold/Contour level (should reflect what works in ChimeraX)… This is before getting into any parameters for the Local Refinement job itself.