Hi all,
Firstly, my apologies for posting off topic during FSC Weekend… ;)
I’d like to understand more about how the mask used during relion_postprocess is applied during Bayesian polishing.
Empirically I've found that a relatively generous mask works best. A tight mask (but which still encompasses the whole protein), like the auto-tightened mask used during final FSC calculation in cryosparc gives good results in post processing, but tracks from polishing look weird and resolution is decreased after refinement. In contrast, a more generous mask gives lower resolution during postprocessing, but good results during polishing.
Running postprocess with a tight mask, and then editing the postprocess star file to point to a more generous mask gives more or less the same results as using a generous mask for postprocessing to start with, indicating that it is the tight mask itself which is responsible for aberrant behavior during polishing, not the FSC or anything else.
Is the mask being applied to the frames during the FCC calculation step, or is it being used in some other manner? What is the best procedure for choosing a mask for postprocessing when one intends to proceed with bayesian polishing?
Cheers
Oli
########################################################################
To unsubscribe from the CCPEM list, click the following link:
https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=CCPEM&A=1
|