unblur & ctffind

Submitted by luwei on Mon, 2015-09-21 02:25

Forums

Unblur & Summovie

Hi Tim,

When doing ctf estimation, should I use motion-corrected and dose-filtered sum or simply the motion-corrected sum?
I have used the former. Because I thought particles from these sums are used for all further processing.
Could you please explain what's difference and which one I should use?

For comparison I tested 10 micrographs using ctffind4 and the results are pasted below.

Thanks,
Wei

motion-corrected and dose-filtered sum:
21861.669922 21284.919922 45.610001
23154.029297 22563.410156 29.959999
24590.570312 23808.759766 29.600000
26114.150391 25574.759766 26.879999
12914.540039 12461.360352 22.200001
14061.429688 13745.669922 3.420000
14531.250000 14230.339844 3.110000
14936.980469 14559.030273 27.680000
15954.759766 15741.849609 20.420000
16312.480469 16169.120117 8.840000

motion-corrected sum
21875.189453 21305.789062 48.119999
23135.980469 22527.740234 30.540001
24599.570312 23973.550781 30.910000
26023.830078 25537.949219 32.619999
12896.870117 12417.959961 22.010000
14052.509766 13747.150391 3.200000
14486.839844 14184.129883 7.410000
14905.620117 14503.230469 29.990000
15973.629883 15728.780273 21.070000
16331.889648 16185.650391 6.850000

timgrant Mon, 2015-09-21 08:34

Hi Wei,

We generally use the non filtered sum for CTF estimation, as the filter is removing signal from the images that can be useful for the CTF estimation, especially if you have carbon etc in your images. In the case where the spectra looks very good for both the filtered and unfiltered images, you can probably get away with using the filtered images, and this saves you having to create the extra file. Looking at the values you posted above, the differences are not really significant and so I doubt it makes much difference in this case.

Tim