Heterogenous Refinement Error that leads to failure

[hsnyder]

@newbie

Should I then export the processed micrographs from live into regular cryosparc and start with blob/ring picking? Did I understand you correctly?

Yes that’s what I’d recommend. You can use the “export exposures” button in the Live UI to export them cleanly. You should see an export job appear in the regular cryosparc UI, which you can then connect to a blob picker job.

[newbie]

Okay, @hsnyder I’ll start with it stat and post any updates here. Thank you for your help so far.

Do the extra details in my previous comment add any information? Is there a way to trace the input of 2d class jobs started in live?

@newbie

• I also just remembered a detail about the 2D class jobs that I forgot to mention that I think is relevant. In live, I extracted my particles a few times, first using the blob picker. Then, I made 2D classes and did template picking, and then I made 2D classes again. I did a few iterations of this using the ring picker once, and all of this was in live.
  Because the picking is all done under the ‘live worker’ job, I’m not sure how I can see which group of picked particles (blob/ring or template, and if template, which one) were used to generate the 2D classes since they all have an input description similar to the image attached below. That’s also kind of why I forgot to mention this point since the individual picking jobs aren’t listed individually in the workspace. I’m not sure if there’s a way to get this information somehow.

[hsnyder]

Hi @newbie, no problem, happy to help. While I agree that the iterative process you’re describing is probably related, I think it could be difficult to trace the exact “route” the particles took through all the jobs, whereas I’m pretty confident that if you just export the exposures and start over from picking, the problem will go away. So I’d prefer to pursue that avenue first if possible.