How can I apply the local-refined particle parameters to the original particles?

Local Refinement
1

In cryosparc, is it possible to use the parameters of the local refinement (use subtracted particles) to recalculate a map using the original particles? Thanks a lot.

Bests,
Xing

2

Nope. But you can convert them to a star file (with csparc2star.py or the upcoming export job) and then reconstruct in Relion using those parameters.

@apunjani I would love it if the cryoSPARC reconstruction routine was exposed as a job, or even better a command line call that didn’t require database entries. Based on ResLog run times the simple reconstruction code in cryoSPARC is orders of magnitude faster than any other program.

2 Likes
3

Unfortunately as I noticed, the particle orders are different in the un-subtracted and subtracted stacks, so I simply cannot use the parameters obtained by the local refinement (using subtracted particles) in relion (apply to un-subtracted particles). Any solution? Thanks a lot.

Bests,
Xing

4

Hmm, that’s the point of the rlnImageOriginalName field, which I don’t think we have available. If we can get rlnMicrographName, rlnCoordinateX, rlnCoordinateY then this:
star.py --merge-fields rlnImageName --merge-source original.star csparc_subtract.star new.star
wil do what you want.

Can you make “csparc_subtract.star” using --copy-micrograph-coordinates original.star or using a passthrough file? (In order to get the micrograph coordinates to use for merging).

5

Thanks, Daniel,

the particles orders and stack names in csparc_subtract.star and original.star are different, and I cannot find information in csparc_subtract.star that could help to point back the subtracted particles to the original particles.

BTW, what are the roles of the passthrough_particles.cs in cryosparc? is this the passthrough file you talk about? Thanks a lot.

Bests,
Xing

6

My merge commands correctly patch particles regardless of the physical order in the file (that is why I wrote them).

The micrograph name and x, y coordinates of course are the same, which is why I suggest merging based on those above.

The passthrough files for some cryoSPARC jobs contain additional information for the particles which is not convenient to place in the refinement metadata file. Primarily some CTF parameters and particle paths. The micrograph coordinates can be obtained (sometimes) from that file, but you can always get them from your original star file as I described above.