Question about the power spectrum in CTF estimation job

CTF Refinement
1

Recently, I was preparing a lecture for the LBMS cryo-EM workshops and found that I don’t really understand this plot. I have two questions.

  • The power spectrum is the squared product of the CTF, structure factor, envelope function, plus background noise. The local minima are used to estimate the background noise. After noise subtraction, the minima should be at the same level (either zero or a baseline value). In CTFFIND4, the amplitude spectrum is used (the square root of the power spectrum), and in GCTF, the log amplitude spectrum is used. In both cases, the minima should be at roughly the same value. Why do the PS minima in this figure not lie at the same value? Why does the PS look like the CTF itself?

  • The CTF in the figure looks like a phase-flipped CTF — all the minima are at zero.
    Why is that?

I discussed this with my colleague but still do not fully understand it. I hope either Oli or the cryoSPARC team can help. (@olibclarke @rwaldo @apunjani )

Thanks.

Liguo

3

Please continue discussion of these questions under the related topic Interpreting plots in CTF estimation job, where these questions have also been asked.