I’m wondering if im correctly using the exposure group utilities job.
Basically, when I collect data with serialEM (3x3 multishots), I add the navigator item into the name of each movie, which looks like : Date_Sample_68-36_0001.tif. 68-36 corresponds to an example of navigator item, with the first number corresponding to the gridsquare.
Then, after achieving my highest resolution with NU-refine, I use Exposure group utilities job to split my output particles according to the navigator item number. I usually end up with 10 to 15 exposure groups corresponding to 10-15 gridsquares of the collection. Next I run 10-15 Global CTF refinement (1 for each exposure group) and Homo/NU-refinement by combining all of the Global CTF refinement.
Is it the way to do it? I didn’t notice a lot of improvement compared to doing Global CTF refinement on the fly during the latest NU-refinement without splitting my particles. I read that this job is useful for image shift group, but how can I discriminate the image shift group in the name of my particles / micrographs ?
Sorry if im not clear but this really confused me. Thanks!
Thanks for your questions, it sounds like you made a good start at splitting into exposure groups to refine Global CTF parameters! Where data is collected with image beam shift you may find it more beneficial to split your exposure groups on the basis of those groups rather than on the basis of grid squares.
As the data were collected using SerialEM, the file naming will depend on how the collection was set up, and it may be best to ask your collection facility which part of your filename contains the image shift group. However, the last number of your movie file name may be the image shift and this is especially likely if all your files end in 0000.tiff to 0008.tiff, so you could try re-running Exposure Group Utilities to split using this part of the file name.
Once you have multiple exposure groups assigned to your particles, some good news for streamlining your processing pipeline is that it is not necessary to run a separate Global CTF Refinement job for each one. As long as the assigned exposure group numbers are different, CryoSPARC will refine them individually within a single Global CTF Refinement job (or during Global CTF Refinement within Homogeneous Refine or NU-Refine).
I hope this has made things a little clearer for you!
Thanks for your detailed answer.
I think I misunderstood the image shift signification, so yes, it sounds clear for me that image shift are represented by 0000 to 0008 at the end of the micrograph name.
When I use Exposure group utilities Job, I split my micrographs like this, I obtain 9 exposure groups.
Do I need to re run Extract From Micrographs job to assign the particles to each exposure group, or the particles are directly associated during exposure group utilities job ?
In the meantime use the “Correspond particles to exposures and enforce consistency of exposure group IDs” option in Exposure group utilities job, and particles seem to be connected to their respective exposure group. Lets see !