Hi we’re collecting data on a Krios with a K3 using 16 bit tiffs, and I’m getting an odd error in motion correction that I haven’t seen before in any of our previous runs with other K3 data (different systems) or other camera on this scope and didn’t know if anyone else knew what this error was:
Traceback (most recent call last):
File “cryosparc_worker/cryosparc_compute/jobs/rtp_workers/run.py”, line 370, in cryosparc_compute.jobs.rtp_workers.run.rtp_worker
File “cryosparc_worker/cryosparc_compute/jobs/rtp_workers/run.py”, line 431, in cryosparc_compute.jobs.rtp_workers.run.process_movie
File “cryosparc_worker/cryosparc_compute/jobs/rtp_workers/run.py”, line 577, in cryosparc_compute.jobs.rtp_workers.run.do_patch_motion
File “cryosparc_worker/cryosparc_compute/jobs/rtp_workers/run.py”, line 582, in cryosparc_compute.jobs.rtp_workers.run.do_patch_motion
File “cryosparc_worker/cryosparc_compute/jobs/motioncorrection/patchmotion.py”, line 255, in cryosparc_compute.jobs.motioncorrection.patchmotion.unbend_motion_correction
File “cryosparc_worker/cryosparc_compute/jobs/motioncorrection/patchmotion.py”, line 302, in cryosparc_compute.jobs.motioncorrection.patchmotion.unbend_motion_correction
AssertionError: Get rid of patches that are too close to edges
Don’t know why I’m getting it now or best way to fix. Thanks for any tips.
I’ve tried changing the number of override knots from 1-4 and get the same message regardless. I switched to the main cryoSPARC and full frame alignment works on the images. What would be a way I can get full frame alignments to be done in cryoSPARC-Live?
Sorry this slipped through the cracks. Are you using an especially high magnification? At extreme microscope settings, the patch-based approach breaks down—each patch doesn’t have enough signal for the motion correction technique to work. As you’ve found, full-frame motion correction is a work-around. Unfortunately there’s currently no way to use full-frame motion correction in cryosparc live. Patch CTF can suffer from the same limitation as patch motion. I don’t have an example dataset that I can use to confirm this, but you could try using the “Override knots” parameters in the CTF job - setting it to a low number may help circumvent the issue.
Edit: I should mention that this limitation is an area that we are researching internally. While I can’t make any specific promises, I suspect that in a future cryosparc release, we will have better support for high magnification. Very low magnification can also cause issues, which is also something that we are looking into.
Thanks @hsnyder we ended up figuring it out. Sorry for my delayed response. While I’m responding to you do you know who I can contact on your team to setup a meeting with my IT team to go over some of the front end / back end interactions? We’re in the process of looking to expand our cryoSPARC environment to a trans NIH program and they have some questions. Do you know who would be best to email? I’ve emailed the firstname.lastname@example.org 3 times now and even email@example.com with no response. Thanks so much for your help if you can.