I combined two sets of down-sample particle stacks (from two different datasets) for a NU-refinment. Then I noticed that the FSC did not drop to 0 at Nyquist frequency. So I checked duplicated items using job “Remove Duplicates”.
Either of stacks shares few duplicated items (<100) while the combined set gives me a number of ~70000. And I am pretty sure that these two stacks has no intersection.
May be the same down-sample file name of batch_xxx_downsample.mrc do the wrong thing?
@joojle Welcome to the forum.
Were movies recorded independently for each dataset, or were the datasets derived from the same movies, but “forked” at some processing stage?
Please can you share
the questionable FSC plot
pixel sizes (on detector, on movie, particle stack) and movie format for each set
Stacks come from two independent sets of movies.
All movies recorded in tif format.
FSC:
Pixel size:
detector & raw movie = 0.52
local motion corrected paritcle = 0.52
motion corrected micrograph & down-sample particle = 1.04
I did make a mistake to import duplicated movies at first. So I had two copies of my first dataset with different uids and I didn’t realize it until I checked all involved job’s parents.
Sorry to bother you.
Joe
