CTF estimation, CTF refinement and beam tilt during acquisition


I have data that have been collected with beam tilt, so I’ve been reading about exposure groups and different threads on this discussion list and wonder what are the current best practices.

From what I understand, patch CTF estimation is good enough to get to a 3.5-4Å resolution. If the data can go to higher resolution and one wants to go into CTF refinement and higher order aberration corrections, then one would go ahead and split the particles (micrographs?) using the exposure utility and go forward on CTF refinements.

Is this the correct way or is it better to start splitting the movies into separate exposure groups right from the start ?

Thank you

Hi Vincent, both will work, it’s just a matter of how easy it is to split the micrographs by beam tilt to create the optics groups and when you want to implement this. For EPU collected mics, it should be easy, but for leginon collected mics, you’ll have to split micrographs by x,y beamtilt a bit more manually. Data from some microscopes will benefit from beamtilt-based optics group refinement and others will not (when the microscope has very good coma free alignment, I think). You could certainly implement the optics groups at import and test refinement with and without optimization. I find a large (0.3A) increase in resolution once sub-3A when using a single group with per-group ctf parameter optimization and don’t find any additional increase when splitting by beam tilt groups or even per-micrograph groups, when coma free alignment is enabled during collection (and working well).