Unblur Guinier plot

Hi all,

I am now using Unblur and Summovie to process my data. I am curious to know what the Guinier plot usually look like with the dose filter applied and noise power restored, is the b-factor going to be smaller than reconstruction without the filter? Or is a linear fit of b-factor agrees well with the experimental map amplitudes decay?
Thanks a lot.

TAT

Hi,

The restore noise power filter should restore the power of the noise back to the level it would have been in the non exposure filtered sum. The power of the signal will be different however, and so the amplitudes will likely be different in the final reconstruction between the exposure filtered and non filtered.

In terms of SNR, one could roughly say that when comparing the sum of all frames to the exposure filtered sum of all frames, the two will have around the same SNR at low spatial frequencies, but the exposure filtered sum will have higher SNR at high spatial frequencies. In this case, the reconstruction would look sharper than the non exposure filtered reconstruction, and thus the apparent b-factor would appear to be smaller.

Comparing the exposure filtered sum of all frames, to a sum of just the optimal number of initial frames, the opposite is true. In this case the SNR would be similar at high spatial frequencies, but higher at low spatial frequencies in the exposure filtered sums. A reconstruction from the exposure filtered sums would look more blurry (due to the increased low resolution power), and thus the apparent b-factor would be higher.

My guess is that the linear b-factor fit will be just as valid as an estimate of the b-factor, since in any case it is just an approximation of the filter caused by many different effects, the exposure filter is simply an additional effect.

Cheers,

Tim