IndexError due to excluded views

[apeck12]

Hello,
I'm running into the following IndexError after tilt-series alignment during tomogram pre-processing. For alignment, I'm using aretomo2 with the DThreshold left at its default value of 0.7. I thought this error might be from dark frames not being handled correctly, but the tomograms that were successfully processed all have dark frames (though it looks like in those cases, the excluded frames are the final tilt images rather than in the middle of the tilt range). I'd appreciate any advice on how to resolve this.
Thanks so much,
Ariana

2025-08-11 11:13:38 [INFO] pyp/align/core.py:75 | Tilt-series alignment took: 9m 6.9s
2025-08-11 11:13:38 [INFO] /opt/pyp/bin/run/pyp:75 | Total time in tomo_swarm 11m 21.4s seconds
2025-08-11 11:13:38 [INFO] /opt/pyp/bin/run/pyp:3586 | Filesystem               Size  Used Avail Use% Mounted on
2025-08-11 11:13:38 [INFO] /opt/pyp/bin/run/pyp:3586 | datapool/namespaces/tmp   26T   15G   26T   1% /tmp/pyp
Traceback (most recent call last):
  File "/opt/pyp/bin/run/pyp", line 4171, in <module>
    tomo_swarm(args.path, args.file, args.debug, args.keep, args.skip)
  File "/opt/pyp/src/pyp/utils/timer.py", line 97, in wrapper_timer
    return func(*args, **kwargs)
  File "/opt/pyp/bin/run/pyp", line 1882, in tomo_swarm
    additional_exclude_views = align.align_tilt_series(name,parameters=parameters,rotation=tilt_axis,excluded_views=exclude_views)
  File "/opt/pyp/src/pyp/utils/timer.py", line 97, in wrapper_timer
    return func(*args, **kwargs)
  File "/opt/pyp/src/pyp/align/core.py", line 5204, in align_tilt_series
    angle = line.split(" ")[3]
IndexError: list index out of range
2025-08-11 11:13:38 [ERROR] /opt/pyp/bin/run/pyp:4176 | nextPYP (tomoswarm) failed

[abartesaghi]

Ariana,

Without seeing the top of the log, I can't say for certain which version you're using, but based on the error message, it appears to be version 0.7.0. That part of the code has been updated in more recent versions, so I recommend upgrading to the latest release (0.7.2) and re-running the pre-processing block. If the issue persists, we'd be happy to take a closer look if you are willing to share an example tilt-series.

Alberto

[apeck12]

Thanks for the rapid diagnosis, Alberto. This is version 0.7.0, so we'll try upgrading to the latest release and rerunning the block.

[apeck12]

Hi Alberto, thanks again for the advice. I had another question about the pre-processing block: is it possible to omit the gold bead detection step? I'm using AreTomo2 for tilt-series alignment on a dataset that has no fiducials. I only see a DErase fiducials box (which is not checked) but no other options in the pre-processing settings that relate to gold beads.

[abartesaghi]

If you're using AreTomo for alignment, nextPYP will skip fiducial-based alignment entirely because these methods are mutually exclusive. However, since alignment and reconstruction are independent steps in nextPYP, you can mix and match approaches. For example, you could use IMOD fiducial-based alignment with AreTomo reconstruction, or use AreTomo for alignment and IMOD for reconstruction. So, even if you aligned a tilt-series that contains gold fiducials with AreTomo, you can still use the fiducial-erasure feature by checking the Erase fiducials box in the Reconstruction tab. If that option is unchecked, fiducials will neither be detected nor erased.

[apeck12]

Thanks for the clarification, Alberto. I used AreTomo for both alignment and reconstruction and left the DErase fiducials box in the reconstruction tab unchecked. So my understanding is that fiducials should not have been detected, but I'm seeing a line indicating that they were in the log file, e.g.:

2025-08-12 10:55:11 [INFO] pyp/detect/core.py:75 | Detecting gold beads took: 7m 52.2s

Did I misunderstand how the settings should be set to make sure fiducials weren't detected? In case it matters, I haven't been able to upgrade my installation yet so am still using v0.7.0.

[abartesaghi]

I see your point. We shouldn't be searching for the gold if we're not going to remove it, especially since that operation takes 8 minutes to run! That said, I checked the code and it's just an inefficiency, the gold won't actually be erased. I will add this to our TODO list, thanks for reporting.

[apeck12]

Thanks for confirming and the speedy response!