Greeting there,
My complex is large so I have to use large box size that stop me using the NU because of memory issue, so I bin2 of my particles; one part always dominant my alignment, so I have to do signal subtraction. By doing these two steps, I got results with medium resolution.
Now I want to try reference based motion correction, my questions are as below:
1, is it possible to do reference motion correction with bin and signal subtraction result?
2, the result is with medium resolution, will it has any benefit?
Thanks and happy holiday!
Hi @yanhezhao,
2, the result is with medium resolution, will it has any benefit?
I do not think RBMC is worth it to run for medium resolution, but it could improve the map slightly. It is more of a polishing step to me. Saying that, what is your GS-FSC at ? Can you resolve secondary structure ? If so there are parts of the map RBMC could improve.
In regards to
1, is it possible to do reference motion correction with bin and signal subtraction result?
Why do you need to do signal subtraction / particle subtraction ? Can you just use a mask and local refinement for your region of interest ?
In general with the 2xbin/Fourier crop of your particles what is the box size (1024=>512) ? Most on here say 600 pix is the max that NU refine will run with. Have your tried running the particle stack in another SPA package to see what resolutions are achieved ?
What was your acquisition pixel size ? In most cases e.g. ~0.82A/pix I do not think 2x bin would hurt too much. Along those lines, if your “medium resolution” is 5-8A more binning should not hurt that much if you assume: Nyquist = 2x binned pixel size => 2x (0.82A/pix)=1.64A you can estimate how much you can bin without loosing resolution or just look for the FSCs to hit Nyquist.
2 Likes