Correction for anisotropic magnification in RelionCor2 or CtfRefine?

[Oliver Clarke]

Hi,

Would it be possible to add the option of correcting for anisotropic magnification to Relion's implementation of MotionCor2? Currently one needs to use the external version of Motioncor2 for this, which then means not being able to take advantage of the Bayesian polishing module.

Alternatively/additionally, perhaps this might be a useful parameter to add to the CtfRefine module, either as a global dataset-level parameter, or refined per mic or per particle like jspr does? (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4935934/)

Cheers Oli

[Oliver Clarke]

(actually re-reading the manual I guess it is possible to use the external MC2 binary so long as it is executed from relion? so maybe the first part of this doesn't apply...)

[Takanori Nakane]

Hi Oli,

You can use UCSF MotionCor2 to apply anisotropic magnification correction. However, Bayesian Polishing deals with un-corrected raw movies without knowledge of anisotropic magnification. This means that the shiny particles revert to an uncorrected state. We understand this is a problem, but are not sure if we can make it by version 3.0.

Best regards,

Takanori Nakane

[Takanori Nakane]

We are working for this now but it will be part of RELION 3.1.