Alignment parameters / metadata 2D class

Dear all,
how can I read out the alignment parameter for each file that has been determined during 2D classification?
Best
Dario

@DarioSB Please can you

  1. describe how you plan to use the parameter
  2. post an example filename whose parameter(s) you wish to read

I want to further process the images in spider and ideally simply apply rotations and shifts to them before classification in spider.
Cryosparc2star gives me rotation and shifts parameter, but when I apply them, the average is not correct

Hi @DarioSB!

If you load the particles after 2D classification into cryosparc-tools, you can find the shifts listed as alignments2D/shift and the rotations listed as alignments2D/pose (these are the angles in radians).

I hope that helps!

1 Like

Thank you for the reply.
So I used the two approaches:

  1. csparc2star.py: convert the particles.cs into a star file, which gives me x,y-shifts and rotations.
    After applying these parameters to the particle stack, the average is just a blob and looks nothing like the average from cryosparc ( i tried first shift, then rotate as well as first rotate, then shift).
  2. cryosparc_tools: I thought that maybe the values from the star file gt mixed up, but the values I get from cryosparc_tools are the same

Any idea why I cannot reproduce the average by simply applying shifts and rotations to the particle stack?

Here you can see two raw images (from a neg stain test-data set) that are off compared to the average.

Or is there ANY way to extract the aligned particles stack?

Hi @DarioSB,

To help identify where your workflow is encountering issues, it might be worth trying the following (if you haven’t already):

  1. select a single class from 2D classification using Select 2D Classes.
  2. using the most recent version of pyem, convert the Select 2D Classes job cs file (with the passthrough cs file) to a star file via csparc2star and check that the necessary columns are present.
  3. perform 2D reconstruction of these particles in RELION with CTF correction for comparison with the CryoSPARC 2D class average.