Output from CTFFIND 4 do not match

Submitted by elena on Mon, 2016-07-04 14:27

Forums

ctffind & ctftilt

Hi Alexis,

I have noticed that outputs from CTFFIND4 sometimes do not match each other. In particular, .txt file shows that ctf rings were fitted successfully only to 34A resolution, but the power spectrum for the same movie indicates much better results (Please see attached). The Nyquist is 5.24A. Would you trust the determined ctf parameters for such movies?

# Columns: #1 - micrograph number; #2 - defocus 1 [Angstroms]; #3 - defocus 2; #4 - azimuth of astigmatism; #5 - additional phase shift [radians]; #6 - cr
oss correlation; #7 - spacing (in Angstroms) up to which CTF rings were fit successfully
1.000000 30851.06 30546.95 -37.49256 0.000000 0.1362694E-01 34.39590

Thanks,
Elena

Alexis Mon, 2016-07-04 16:01

Yes, I would. The reported "CTF resolution" of good fit is a noisy estimate. Probably in this case it dipped below the threshold at 35A before going above it again and the true value should probably closer to 10A. You can check this easily by plotting the diagnostic 1D curves using the script provided with ctffind4.