Unable to perform refinement after particle subtraction

[Leotrosa]

Hi all, In Relion 4, I made a mask on the region of interest of my map and performed a subtract job, centering images on mask after subtraction. Then I try to run a refinement using the subracted particles, but the same error appears:

---

ERROR: No orientation was found as better than any other.

A particle image was compared to the reference and resulted in all-zero weights (for all orientations). This should not happen, unless your data has very special characteristics. This has historically happened for some lower-precision calculations, but multiple fallbacks have since been implemented. Please report this error to the relion developers at github.com/3dem/relion/issues

---

I tried different strategies: -Using the original map as a reference with the reference mask -Doing a new initial model (this works ok) and then using this model for a refinement (gives same error) -Doing a 3D classification to pre-align particles (this works fne) before doing a refinement (gives same error)

Environment:

• RELION version 4.0

Dataset:

• Box size: 500 pix
• Pixel size: 0.845 Å/px
• Number of particles: 76 000
• Description: hexameric protein with one rigid domain and one flexible atpase domain - trying to get the flexible domain

Job options:

• Type of job: [Refine3D]
• Number of MPI processes: 9
• Number of threads: 2
• Full command (see note.txt in the job directory):

  ++++ with the following command(s): 
  `which relion_refine_mpi` --o Refine3D/job264/run --auto_refine --split_random_halves --i Class3D/job262/run_it025_data.star --ref Class3D/job262/run_it025_class001.mrc --ini_high 40 --dont_combine_weights_via_disc --scratch_dir /scratch --pool 200 --pad 2  --skip_gridding  --ctf --particle_diameter 180 --flatten_solvent --zero_mask --solvent_mask Import/job263/mask_ATPase_subtract263.mrc --solvent_correct_fsc  --oversampling 1 --healpix_order 2 --auto_local_healpix_order 4 --offset_range 5 --offset_step 2 --sym C1 --low_resol_join_halves 40 --norm --scale  --j 2 --gpu ""  --pipeline_control Refine3D/job264/

=================

[biochem-fan]

-Doing a new initial model (this works ok) and then using this model for a refinement (gives same error)

Are you sure your mask and the new initial model are consistently aligned?

[Leotrosa]

With the refinement comming from the initial model I did not even use a mask, and it crashed anyway with the refinement after the 3D classification, I double checked before launching and they are well aligned, yes

[biochem-fan]

What happens if you make a reference by relion_reconstruct on the subtracted particles?

[Leotrosa]

Hi just tried doing a relion_reconstruct on the subsract particles (which works ok) but the refine crashes with the same error

[Leotrosa]

However, checking the relion_reconstruct result, it resulted in a sphere full of noise and no density inside. Does that mean the problem is potentially back in the subtraction job?

[biochem-fan]

You have to low pass filter the reconstructed map (or reconstruct half sets separately and run postprocess) before visually inspecting it.

[Leotrosa]

Thanks, the reconstruct map looks ok after lowpass filtering