Select particles by latent coords

3D Flexible Refinement
1

dear cryosparc team,

I’ve been obsessed the powerful 3dflex and got many interesting results. One thing I do want to try is to classify particles based on the latent coordinate especially if the histgram looks like multiple gaussian peaks. I found select_by_latent_coords.ipynb · GitHub
pretty much solved this problem except I don’t know how to write the selected particles as an output. say after saving the selected particles as a new pandas dataframe, how do I save this new dataframe as an external result? I guess something like
project.save_external_result(
workspace_uid=“W2”,
dataset=updated_particles,
type=“particle”,
name=“recentered_particles”,
slots=[“location”, “alignments2D”],
passthrough=(job.uid, “particles_selected”),
title=“Recentered particles”,
)
in the recenter particle tutorial. Can someone help me on this?

2

Hi @hwangab! You basically had it right, you just select the passthrough and the slots. I’ve update the example to include the command in the final chunk: https://github.com/cryoem-uoft/cryosparc-examples/blob/main/select_by_latent_coords.ipynb