Clear certain motion correction results

Would clearing Motion Correction clear the non-dose-weighted micrographs? Would that be added as an option?

As Oli and I have asked in the past, keeping two copies of every micrograph (particularly when using super res data) makes the size of the motion corrected data double what it needs to be.

That said, as I/we were testing at the time, patch CTF performs poorly on dose-weighted micrographs so going back to redo CTF estimation after this is a little troublesome (CTFFIND is fine, though…)

@rbs_sci Do I get this right: You are seeking fine-grained control over whether to clear

  • dose-weighted
  • non-dose-weighted or
  • all

motion-correction output?
[I moved your post to its own topic. Please let me know if the topic heading does not match the “spirit” of your suggestion.]

1 Like

Hi @wtempel, it would be great to be able to clear the non-dose-weighted mics, as they are only needed (or maybe not?) for CTF estimation, so once that is done they are just taking up space and effectively doubling the required disk space.

I don’t think it is necessary to clear the dose-weighted mics (in that case one might as well just delete the job).

1 Like

Thanks for moving this to it’s own thread, @wtempel. :slight_smile:

As @olibclarke says, control of non-dose-weighted micrograph deletion. Dose weighted we will (usually) always need.

Although actually it might be useful to remove “failed” dose-weighted micrographs as well; e.g. micrographs which fail CTF, or are too contaminated or possibly melted during exposure acquisition. Shortly after we commissioned our Titan, we had a grid where the ice was razor thin - a statistically significant proportion of exposures vaporised during acquisition, even when I cut the total dose down. Looking back it is actually rather amusing because too thin ice is not usually a problem with many of the samples we study. :smiley:

Seconding this request for an option to delete non-DW mics from motioncorrection jobs as this could quickly free up many hundred GB of space for some datasets.