Patch-based CTF estimation automatically estimates defocus variation for tilted, bent, deformed samples and is accurate for all particle sizes and types including flexible and membrane proteins. This is accomplished by a fast GPU implementation that usually takes 1 - 2 seconds per micrograph. No prior knowledge about particle location is needed.
Variable defocus example plot. Vertical axis shows defocus (A) as a function of position.
Patch CTF compares the computed CTF to the observed power spectrum of the input micrographs, and generates a plot illustrating how well they match. In the example plot below, notice that the plot title reports "FIT 3.327 A" - this means that the computed CTF matches the observed power spectrum up to a resolution of 3.327 angstroms. The cyan line in the plot shows the cross correlation of the observed spectrum and the computed CTF. CryoSPARC uses 0.3 as the cross-correlation threshold for a good fit - the vertical green line in the plot corresponds to the frequency at which the fit drops below this threshold.
CTF fit example plot. Black: observed power spectrum. Red: calculated CTF. Cyan: cross-correlation (fit).
Note that when patch CTF is run, the output micrographs will contain information about their average defocus and the defocus landscape. This information is automatically used when particles are extracted, to assign each particle a local defocus value from its position on the landscape.
Patch CTF also outputs other diagnostic information with the micrograph, including and estimate of ice thickness, and the CTF fit score. These are used in exposure curation to separate good and bad images.
For a theoretical overview of the CTF and information about the CTF model used in cryoSPARC see the page linked below.