[3dem] Query regarding particle loss in RELION polishing

[Haaris Ahsan Safdari]

Hi all,

I ran a polishing job in RELION 5.0 (commit-4e57a4) using the optimal
parameters obtained from training. My input consisted of 585,635 unbinned
particles (full box size). The dataset was acquired on a K3 detector in
TIFF format with 40 frames per movie.

However, upon running polishing, I noticed in the job log that several
movies in this dataset contain fewer than 40 frames. Since this dataset was
acquired by someone else, I am unsure why this is the case. The log
includes warnings such as:


- The following micrographs do not contain 40 frames. Particles in them
will be ignored:
  Movies/mic_Dec22_22.17.24.tif
  Movies/mic_Dec22_22.17.49.tif
  Movies/mic_Dec22_22.17.53.tif
  Movies/mic_Dec22_22.18.01.tif
  Movies/mic_Dec22_22.18.06.tif
  ...
WARNING: 215,753 particles in optics group 1 were removed.


Although the polishing job completed successfully, 215,753 particles were
ignored. As a result, when running a 3D refinement job on the polished
particles, only 369,882 particles (instead of 585,635) are copied to the
scratch directory.

I am concerned that this significant particle loss may impact the
resolution gain I would expect from polishing. Could anyone suggest a
strategy to retain these 215,753 particles while ensuring proper
per-particle motion correction?

Best regards,
Haaris Safdari
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.ncmir.ucsd.edu/pipermail/3dem/attachments/20250211/26f978b0/attachment.html>