Weird relative ice thickness plot while using dose weighted images

Features and Functionality
1

Hi

I have a previously motion corrected dataset and all the plots were fine when I used non-dose weighted images for both CTF estimation and particle extraction.

P14_J5_power_spectrum_for_006955623134459049834_000057481574545983473_foilhole_30337_data_47591_47593_20220101_080404_fractions_patch_aligned
P14_J5_power_spectrum_for_006955623134459049834_000057481574545983473_foilhole_30337_data_47591_47593_20220101_080404_fractions_patch_aligned1581×506 47.8 KB

Later on, I read that it is better to use dose-weighted images for particle extraction (Dose weighted and non-dose weighted) As I didn’t have motion correction done in my cryosparc instance, I decided to try both patch CTF estimation and particle extraction with dose-weighted images. In the plots I got from the CTF estimation job, the relative ice thickness plot seems a little odd to me although the diagnostic plot with PS,CTF, and FIT is good. The ice ring location seems fine. But the curves are shooting up.

P14_J95_power_spectrum_for_016007911467120795840_000136460248857936181_foilhole_24037_data_47591_47593_20220101_045110_fractions_patch_aligned_doseweighted
P14_J95_power_spectrum_for_016007911467120795840_000136460248857936181_foilhole_24037_data_47591_47593_20220101_045110_fractions_patch_aligned_doseweighted1581×506 30.6 KB

I am having trouble understanding what’s going on. Any help is appreciated

2

@iptom In cryoSPARC, references to dose-weighted and non-dose-weighted micrographs are usually passed on to downstream processing alongside each other. Downstream processing jobs are expected to select the more appropriate input for the given task without explicit user intervention, non-dose-weighted input in case of CTF estimation.
While it may be “possible” to run a CTF estimation job in cryoSPARC with dose-weighted inputs, we do not recommend dose-weighted input for this job type.

1 Like
3

Thanks for the kind reply @wtempel