Question about using cluster3D for heterogeneous class averages

[JulianMoebius]

Hi! I have a heterogeneous dataset with about 200 class averages coming from multiple conformations, and I wanted to try the multi-volume reconstruction method (cluster3D). However, when I run it, I get the following error:

ChildProcessError: `simple_exec prg=cluster3D nstates=2 pgrp=c1 hp=250 lp=10 lpstop=10 lplim_crit=0.3 lpstart=20 msk=44 nparts=2 nthr=0 projfile=/home/julian/Documents/SIMPLE/SIMPLE-3.0.0/build/test_capsid_project_folder/test_capsid/1_import_cavgs/test_capsid.simple` failed with code 1

I’m not sure whether I’m using the method correctly or if I’m misunderstanding the purpose of cluster3D. Could you clarify how it is intended to be used, or what might be causing this error?

Thanks in advance!

[hael]

Hi again,

I assume that you are using the latest release branch, not the dev branch (master). Cluster3D is not maintained any more unfortunately and we don't typically recommend doing heterogeneity analysis on class averages. If you pull the development branch (master) there's are two programs abinitio3D and multivol_assign that can do heterogeneity analysis as described in our recent IUCR paper, but you will need particles, not class averages. I am in the middle of some pretty serious refactoring of the code at the moment, but I can help you more directly in a few days. Email me directly: hans.elmlund@nih.gov

Best

[JulianMoebius]

Thanks a lot for the quick response! However, my goal is to work with class averages, so I’d prefer a solution that doesn’t require going back to the particles.

[hael]

200 class averages sound like a too crude discretization to get any meaningful heterogeneity analysis done. Good luck!

…

________________________________ From: JulianMoebius ***@***.***> Sent: 02 December 2025 05:15 To: hael/SIMPLE ***@***.***> Cc: Elmlund, Hans (NIH/NCI) [E] ***@***.***>; Comment ***@***.***> Subject: [EXTERNAL] Re: [hael/SIMPLE] Question about using cluster3D for heterogeneous class averages (Issue #20) [https://avatars.githubusercontent.com/u/104631304?s=20&v=4]JulianMoebius left a comment (hael/SIMPLE#20)<#20 (comment)> Thanks a lot for the quick response! However, my goal is to work with class averages, so I’d prefer a solution that doesn’t require going back to the particles. — Reply to this email directly, view it on GitHub<#20 (comment)>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/AAJENUQ53GQ74J2OGILZPBT37VRENAVCNFSM6AAAAACNRITAGOVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZTMMBRGI3DCMBWGU>. You are receiving this because you commented.Message ID: ***@***.***> CAUTION: This email originated from outside of the organization. Do not click links or open attachments unless you recognize the sender and are confident the content is safe.