Histograms rather than (or as well as) micrograph plots for CtfRefine?

[Oliver Clarke]

Hi,

Currently CtfRefine generates per micrograph plots, where each particle is colored by defocus and the defocus spread is printed.

I use these to judge if CtfRefine has worked - usually, when the defocus spread is comparable to the search range, things have gone pear shaped (assuming that the search range is much greater than the expected thickness of the ice, in a non-tilted case).

These plots are useful, but I think a histogram, with the original (global) defocus marked would be a useful additional diagnostic for this purpose, and would also maybe allow the user to see whether the global defocus might be systematically wrong for particular micrographs (if the per particle defocus values are all clustered at one end of the search range).

Attached is a not-very-pretty example generated for a micrograph from a current run (global defocus marked in magenta), along with the matching logfile entry.

Cheers Oli