Symmetry expanded particles yields artificially higher resolution

3D Refinement
1

Dear All,

I tried symmetry expansion on my dataset followed by refinement in C1. The resulting map FSC looked too good to be true (I got ~3 Angs according to FSC cut-off of 0.143), and upon closer inspection of the map I could see that the resolution was much worse than the non symmetry expanded particles refined in the symmetry of the protein (~3.9Angs).

Any ideas why this is happening? Am I doing something wrong?

When I repeated the run without a mask from a previous refinement that seemed fine, the expanded particles refinement hits nyquist! but again looks terrible.

Best,
Omid

2

You should only perform local refinement with sym expanded particles - same in relion - global search you will superimpose the different sym expanded copies

Oli

1 Like
3

Hi @olibclarke,

Yeah that makes a lot of sense. I figured I was not doing something right. I should have used a mask for the monomer which is what I wanted.

Thanks,
Omid