Poor CC values in

Submitted by rkman2 on Wed, 2016-01-13 11:03

Forums

ctffind & ctftilt

Hi,

I'm getting very poor CC values (in both the txt files) when using ctffind-4.0.17. The CC values for ctffind-4.0.7 are much better; however, the generated power spectrum from ctffind-4.0.17 matches the micrograph much better than ctffind-4.0.7. I would appreciate any suggestions.

Alexis Wed, 2016-01-13 11:48

Hi there,

At version 4.0.16, the background subtraction algorithm was amended (to be more similar to the one in ctffind3). This led to improved reliability of fits in test datasets and it seems in your case it let to a better fit also, which I'm glad to hear.

It could be that because of this change, the normalization of the CC score was affected, I am not sure. In any case, I would not pay much/any attention to the absolute value of the CC score, especially when comparing ctffind runs with different versions and/or parameters.

Hope this is helpful
Alexis