Extract from mics - larger dataset gives lower achievable resolution

Hi all,

I first extracted at 1260 px / 560 px (extraction/fourier) and after processing with 300 micrographs I can get to a resolution of 2.5 A (data doesn’t go that far but FSC shows that limit). When I scale up to 3000 mics, the FSC tops at 9.5 A. Is this correct? Or have I made some mistake elsewhere?
Matt

Sounds like something has gone wrong or there is some point of confusion. Perhaps you could post pictures of the FSCs?

This is from hetero refinement. But this is after using all particles.

image

This is from homo refine and 10% of micrographs.

Extract was done in each case with the above box sizes but in two jobs, 1 with 300 mics one with 3K mics. Does hetero truncate resolution range?

Oh, i see by default the box size is 128 px…

1 Like

Yes, you got it - in hetero the calculations are run by default in a 128px box for speed

1 Like