Hello forum,
So I’ve been working with a EM dataset.
I’ve obtained an map with 5.6 resolution (178 mb in size) but when I open it in Coot it doesn’t have details.
This in my processing steps:
Blob picker - inspect picks - Extract particles (512/128)- 2D classification - Select 2D classes- Template picker - inspect picks- extract particles (360-no binning) - 2D classification - select 2D classes - Ab-initio (4 classes) - Heterogenous ref (4 Classes) - Homo ref (2 classes, 137K particles) - NU-ref - 3D classification.
I was wondering if there is a appropriate size for a more detailed output map.
Thank you!
Hello,
Can you elaborate by what you mean by “details”, can you see something or nothing at all? At 5.6 you will not see side chains, at best you might see some secondary features (helices / sheets), but primarily tertiary structure. If you do see something, you might be limited by number of particles, orientations, sample size and happiness, collection parameters.
Best
Thank you for reply.
Yes. By details I mean mostly side chains.
So I need more particles and better resolution to see those details.
A lot will depend on your grids. A good sample (stable biochemical prep, etc) and good grid will get high resolution without many particles, but a bad grid and/or biochemistry will not reach high resolution no matter how much data you collect and many particles you throw at a reconstruction.
If you can share some example images (a micrograph or two, a picked micrograph, 2D classes), perhaps we can offer some insight into whether your current data might achieve your target, or whether you need to go back and optimise. 
I would not expect to see side chains (bumps off the main chain) above ~4A. The medium resolution can still be helpful if there are alpha helices and other secondary structure to follow.
can play with the slider at the bottom of
https://cryoem101.org/chapter-5/
1 Like