crystalline ice flag?

Submitted by oliverbclarke on Sat, 2015-09-12 15:21

Forums

ctffind & ctftilt

Hi all,

CTFFIND4 has been working wonderfully for us, many thanks for its development and support!

I was just wondering whether it might be worth at some point adding in an option to flag (or discard) images with crystalline ice? This is typically very obvious in the power spectrum, and I guess all that info is in the *_avrot.txt files... this would be very handy when analyzing large datasets for directly getting rid of these micrographs without having to visually inspect them (or at least to flag them for inspection).

I can probably write a script to do this myself, but I'm not exactly sure how to parse the output - is the value in row 3 (1D rotational average of spectrum) normalized somehow? and if so is the raw data anywhere? I would like to be able to flag images based on anomalously high values in a particular range of spatial frequencies.

Cheers,
Oliver.

Alexis Sat, 2015-09-12 16:10

Hi Oliver,

Unfortunately, the values in row 3 have been re-normalised for plotting purposes as described in the paper. I have made a note to add non-normalized output in the next version, but I'm not sure when this will happen. So in the meantime, the only way you could do this would be by hacking the code and recompiling.
If you wanted to do this, you could easily turn off normalization of the 1D avrot by changing line 3127 of core/images_core.f90 from:

logical, parameter :: rescale_average = .true.

to:

logical, parameter :: rescale_average = .false.

That way, you'd get non-normalized output and you could write a script to parse the avrot files as you suggest. The disadvantage is that the diagnostic plots will probably not look as good any more. I think other aspects (the estimate of highest resolution with good fit) would still work fine.

Hacking to have both normalized and non-normalized 1D profile output would require significantly more work.

Cheers,
Alexis