Frame Alignment & Metadata Recognition

[AriaGarrett]

I'm running into an issue with Motioncorr when I check 'read metadata from mdoc files.'

An index error is happening after tiltxcorr runs successfully but only on a single section, so NextPYP gets an invalid .prexf file.

I believe nextPYP imports the individual tilt movies as independent tilt-series instead of assembling all tilts into a single .st stack.

2025-10-28 11:00:01 [INFO] Combine into one tilt-series took: 0.8s
40	2025-10-28 11:00:01 [INFO] Original tilt-series dimensions = [ 11520, 8184, 1 ]
41	2025-10-28 11:00:06 [INFO] Tilt-series pre-processing took: 4.9s
42	2025-10-28 11:00:06 [INFO] Transformed tilt-series dimensions = [ 11776, 11776, 1 ]

When I uncheck 'read metadata from mdoc files' and give it the pattern TILTSERIES_SCANORD_ANGLE.tif I don't get the error, but I'm worried I'm losing the metadata related to things like inital dose, dose rate, etc. during processing. Or maybe the metadata isn't lost in the processing and I can use this work around.

Any thoughts? Thank you!

[abartesaghi]

The logs suggest that you might be using the *.mdoc files associated with individual frames rather than the ones corresponding to each tilt series. SerialEM typically generates both, but nextPYP only requires a single *.mdoc file per tilt series.

If using the movie pattern works for you (and assuming the file names correctly reflect the scanning order and tilt angle), the only additional metadata you’ll need is the dose per tilt, which can be entered in the Dose rate field under Microscope parameters in the data import block. Note that this field should contain the dose per tilt (in e⁻/Ų), not the dose rate, despite what the parameter description suggests.