Do particles get re-centered for local motion?

I took my initial set of binned particles to homogeneous refinement, then straight to local motion. Do I need to re-extract, refine, and then send to local motion or do particles get re-centered at the local motion step?

Hi @user123, you can take your initial set of particles to homogeneous refinement and then straight to local motion as you said. As long as you correctly set the box size in local motion correction (which itself performs extraction from the raw movie frames) you will end up with correct unbinned particles that can then be taken forward to refinement again.

but they won’t be re-centered, right? it won’t take the offsets into account?

1 Like

Thanks, Saara. This did work but then my resolution was slightly worse than the unbinned dataset without local motion so I was wondering if it’s worth re-extracting those binned particle to re-center them and then send them to local motion.

@olibclarke they will not be re-centered, no - offsets are not used.
@user123 re-centering could have some effect in this case, unfortunately there isn’t any easy way to try this within cryoSPARC at the moment

1 Like

Good to know. Thank you.

@user123 you could re-extract in Relion with recentering, and then re-import into cryosparc prior to local motion (and re-extract in cryosparc) - that way you can test if re-centering helps

Also in my experience local motion will only help if there is a fairly decent number of good particles per micrograph - if particles are sparse it often either makes things worse or doesn’t improve


1 Like

or use --rencenter from pyem package