Does ab-initio refinement support helical particles?

[jplopez]

Hi everyone,

I’m testing the new ab-initio refinement job in cryoSPARC v5 using particles with helical symmetry. The only parameters I’ve changed from the default are Volume window mode (set to Cylindrical window) and Window diameter (inner). However, the job fails with the following error: KeyError: ‘filament/filament_pose’.

Does this mean that thist job does not support helical particles yet, or am I missing some required input or setting?

Thanks a lot for any insight!

Jorge

[mmclean]

Hi @jplopez,

Thanks for your post. We will investigate this internally and see if we can reproduce it. In the meantime, could you let us know how many particles the job was run with, as well as at what point the job failed? (Stating after how many iterations the job failed would suffice)

Thanks,

Michael

[jplopez]

Hi @mmclean

Thanks for looking into this. The job was run with 147,099 particles. It does not reach any refinement iterations, it fails immediately after the initialization step, right after generating the random initial density. Interestingly, the same particle stack runs without issues in the standard Ab-Initio Reconstruction job, which suggests the problem may be specific to the new Ab-Initio Refinement implementation.

Please let me know if you need the full log or any additional details.

Best regards,
Jorge

[mmclean]

Dear @jplopez,

Thanks for bringing this to our attention. There indeed is a bug in homogeneous ab-initio refinement when using helical particles. As a workaround, you can disconnect the “filament” slot associated with the input particles to a homo ab-initio refinement job; please see the GIF below.

If you later run any downstream jobs that use the “filament” slot (e.g. helical refinement), you can at that point reconnect the filament slot any job upstream of the homo ab-initio refinement job.

Best, Michael

[jplopez]

Thanks @mmclean, this job now runs without issues.

Best,

Jorge