Relion5 3D refinement

42 views

Skip to first unread message

Jorgo Pata

unread,

Sep 21, 2026, 3:07:12 PM (4 days ago) Sep 21

to Warp

Hi everyone,

I am working on a membrane protein using a Warp → MissAlignment- RELION 5 STA workflow, and I am seeing some maybe unusual behavior during 3D refinement. I would appreciate some input on whether this is expected behavior or whether there are particular things I should check in the subtomogram extraction/alignment workflow.

The subtomograms were extracted with Warp and are subsequently refined in RELION 5. Box size is 128x128x128, 4 Å/pixel. It hit Nyquist at bin 4 (15.7A), before going for bin2 at the current step.

I performed 3D classification on the dataset and obtain a class with substantially improved density, reaching 12 Å . However, during the refinements, the first iterations stall at 28A-30A, before an improvement during the last iteration (up to 17A or so, but only in this iteration). Is this an expected behaviour for Relion? I also run initial models with the bin2 particles and a 3D class without alignments, the volumes look fine. I would appreciate any advices or ideas.

Reply all

Reply to author

Forward

0 new messages