Hi guys, thank you for the tip, it ended up fixing this problem (3D Flex Reconstruc fails at iteration 0 - #5 by Flow) that I had and thought was attributed to an incompatibility due to an old version of CentOS.
I managed to run this procedure on a couple of dataset without any problem and run all the 3D Flex jobs with downsampled particles.
However, I have a particular dataset where I encounter this error message now when trying to run the downsample particle job:
[2023-04-19 9:28:06.43]
[CPU: 260.7 MB Avail:1947.53 GB]
Traceback (most recent call last):
File "cryosparc_master/cryosparc_compute/run.py", line 96, in cryosparc_compute.run.main
File "/scicore/home/engel0006/GROUP/pool-engel/soft/cryosparc/cryosparc_worker/cryosparc_compute/jobs/extract/run.py", line 893, in run_downsample_particles
assert n.allclose(psize_align_2D, particles_dset['alignments2D/psize_A'], atol=1e-4), "Particles must all have the same alignment pixel size. If multiple particle stacks were input into the job, please instead run separate downsample jobs for each particle stack."
AssertionError: Particles must all have the same alignment pixel size. If multiple particle stacks were input into the job, please instead run separate downsample jobs for each particle stack.
The particles should all have the same alignment pixel size so I don’t really understand.
Originally particles come from two sets of import movies, but from the same microscope with all the same parameters and even from the same data acquisition. The movies were just sorted post acquisition based on astigmatism.
On the other dataset where 3DFlex worked, the particles even come from different data collection, so i’m a bit clueless.
To fix this, I tried to run a Restack job on the particle output from 3D Flex Data Prep and then tried to run Downsample again, but it did not work either.
Any idea on how to fix that?