Local Refinement good, but not after PS

Local Refinement
1

I am working on a protein with a floppy domain so I wanted to use some kind of focus classification. I’ve made a few masks and played with their parameters. I got a nice looking 3.63 map from a local refinement job, but I wanted to see if I could improve it with particle subtraction.

particle subtraction seems to work well, as the reconstruction only job shows that I have isolated my domain of interest. But then when I run the local refinement (with the same mask that gives the 3.63 map) on the subtracted particles, I get low (7-8A resolution) maps.

Is this more likely an issue of my particle subtraction job or the local refinement?

2

This workflow works very well all the time for me. It seems likely there is a mistake like recentering or Z-flip or wrong inputs or something weird. Double check the logs of local refine, is a mask correctly covering the mass? Remember that subtraction and refinement use opposite masks; former of the deleted region and latter of the remaining focus region.