I’m currently using cryoSPARC and Relion 5.0 to process my dataset. On Relion 5.0. I’ve followed the typical workflow or importing my micrographs, doing motion correction, followed by CTF estimation, then autopicking and finally particle extraction.
I’m inspecting my extracted particles and I’ve noticed many of them looking as such (see attached image)
and I don’t think I want to take these forward with 2D classification. Has anyone ever seen anything like this?
I believe the ones with “barcode” (top left particle) is an artefact from extracted particles are the edge of the micrograph and the ones with the “fringes” (top right) is an artefact from the edge of the beam. Is there any way to avoid getting these bad extracted particles, by parameter changes or subset selection of the good particles? Please note: I’ve extracted about 2 million particles.
This is the CryoSPARC forum. Best bet for RELION questions is ask on CCP-EM mailing list.
Now that’s said…
You should dump micrographs with a beam edge encroaching on the sensor anyway.
Barcode edges are common in RELION using the blob picker, see them all the time. The LoG picker has an obsession with them. They usually get removed during 2D classification if the spherical mask is wide enough.
Otherwise, just select good classes and re-pick using the template picker.
Or use cryolo.
Or run a Select job, sort by CoordX, select particles without the barcode edge (select all particles above, select all particles below and invert selection are your friends here), repeat for CoordY, feed selected particles to 2D classification. Or you can do the same in the commandline if you’re comfortable there, but the GUI makes the workflow choices a little clearer.
Best bet for RELION questions is ask on