Feature request: 'scipion - alignment assign'

[JuhaHuiskonen]

When one extracts coordinates ('scipion - extract coordinates') after an alignment there is an option to apply the shifts. I often use this to re-extract particles that are better centered. However, when this is done, alignment information gets lots. Alignment parameters (shifts and angles) can be assigned back to the re-extracted particles with 'scipion - alignment assign'. However, if shifts were applied when creating coordinates for re-extraction, this protocol does not make sense, as it assigns those shifts again (when they in fact should be 0).

I suggest that 'scipion - alignment assign' would have the following options to consider for alignment parameters: ['all', 'only shifts', 'only angles']. In my case I would like to apply only the angles.

Furthermore, when new particles are extracted with updated shifts, this considers only the integer portion of the shift. 'scipion - alignment assign' could also have an additional option 'only sub-pixel shifts' to take this into account.

[pconesa]

Thank you, @JuhaHuiskonen. Seems flexible.

[JuhaHuiskonen]

Any news on when the options ['all', 'only shifts', 'only angles'] can be added to 'scipion - alignment assign'?

I think a lot of users need this as 'only angles' is required every time particles are re-extracted with shifts which is a common thing to do.

[delarosatrevin]

Hi Juha, thanks a lot for the remainder! I haven't find the time to work on this. Some users here in Stockholm has been hit by this issue as well. Next week he have a CryoEM course here, but next one I will take a look at it.

[delarosatrevin]

Dear @JuhaHuiskonen
I'm sorry for taking so long for looking into this. I have implemented the different options you propose for the assignment in this protocol. Right now we will not make a release with this change, since we are in a big refactoring process.
You can find the difference path here:
https://github.com/delarosatrevin/scipion/commit/a1048f881359c9e46ba97ab51126154035ad7247.patch
Could you apply the patch and test it? I have done some tests but would be great if you can double-check.