aligned frames

Submitted by jing on Thu, 2016-07-28 10:00

Forums

Unblur & Summovie

Hi Tim,
When I set "apply dose filter" to "yes" in unblur, is the aligned frames also dose-filtered or simply with the shifts applied?
Thanks a lot,
Jing

timgrant Thu, 2016-07-28 10:03

Hi,

The output frames should be dose filtered with the shifts applied. They are probably not useful for anything as the filters get quite crazy.

Thanks,

Tim

jing Thu, 2016-07-28 10:26

Hi Tim,
Thanks for the clarification!
I would like to use the "frame motion correction" in the EMspring helical reconstruction suite. The program realigns the helix segments from individual frames of the movie stacks but there's no dose-filter applied. Do you think it's helpful to use the dose-filtered frames so that more frames can be included for the final reconstruction or is it better to provide aligned frames without dose-filter but using only the first 10 frames (total 10 e/A2 ish)?
Thanks a lot!!
Jing

timgrant Thu, 2016-07-28 10:32

Hi Jing,

It may work with the dose filtered frames, but it may also cause crazy things to happen, it depends on exactly how spring does the alignment. If it was me, I would probably try both ways and see what gives the best result. If you're trying to save time I would probably go for using the first 10 frames.

If you do try with the dose filter - let us know if it works!

Thanks,

Tim

jing Fri, 2016-07-29 02:56

Hi Tim,
So far I've only tested the dose-filtered frames. With individual frame alignment, I got randomly flipped filament polarity probably due to the low signal in each frame (1e/A2). spring also allows grouping several frames and use the average for alignment (4 frames in my test). It fixed the flipping polarity problem but then I got a FSC curve that falls and stays at 0.2 above ~5A resolution. Do you think it's some kind of masking from unblur? I"m going to try the unfiltered frames and see if that makes a difference.
Thanks for your advice!
Jing