Hi,
I recently begun using Cryosparc 4.4.1, and tried polishing my map using reference based motion correction.
I was able to complete the job successfully.
However, when I used the outputted polished particles as input for a homogenous/non-uniform refinement job, although the nominal resolution improved nicely which was encouraging, I got unusually heavy maps (1.7 GB per map).
When trying to visualize these maps with Chimera, I was not able to see the complete map but only saw slices of z-sections. I’m not sure if this was because the file was too heavy for Chimera or some other problem, but Cryosparc viewer and Coot were able to display the complete map without a problem (though very slowly because of the large file).
I, however, obviously need to be able to visualize the map with Chimera for fitting, masking and visualization purposes…
Any idea why this happened? Is this considered normal? How could I make the map less heavy, and how could I visualize it in Chimera?
The particle stack was resampled to 8K pixel sampling if imported with upsampling 2.
I’ve not checked whether RBMC auto-resamples to original data if input is Fourier cropped, someone else can chime in if that is the case.
The Chimera issue is, as you surmise, related to Chimera defaulting to a visualisation scheme which cuts down graphical load. In the Volume Viewer, change from “Plane” to “Surface” (I forget the exact name, I usually load a settings file with the defaults adjusted to avoid this issue)
After RBMC, the map will be at unbinned pixel size. That’s why it is bigger if you use Fourier crop before. If you want to reduce the size, just use Volume Tool to reduce to the pixel size without affect the resolution of the map.
Reference based motion correction will output particles that are at the same (super)resolution as your original movie.
Fourier cropping needs to be applied at the RBMC level, just as you would do at the motioncorrection level.
The map is bigger because the particles are now at the same A/pixel as the unbinned movies.
You might want to downsample your particle stack as well, unless you really need the super resolution.
