I have a dataset where my complex is about 150 Å wide and 250 Å tall. In many micrographs the particles appear stacked end to end. I routinely see single particles but also stacks of two, three, four, five or six particles. These are not true filaments and there is no helical symmetry. They are simply several copies of the particle sitting on top of one another.
I would like to do three things:
1.Separate particles based on how many units are stacked together
2.Estimate how many of each stack length are present
3.Get low resolution 3D maps for each group
For anyone who has handled this situation before, what is the recommended approach in cryoSPARC?
I have included some 2D classes of the stacked particles and some single particles
The simple answer is (at least for me): I’d dump any aggregated particles completely. The 2D classes for the non-stuck particles look very nice - just work with them and see how far you get.
If you really want to work with the aggregated complexes, feed the entirety of the extracted particles to ab initio and into 10-30 classes, and then pass the whole lot to heterogeneous refinement, then work from there.
If you do see a (highly) repeatable interaction between the aggregated complexes across multiple higher resolution classes, discuss with PI and/or collaborators about whether or not it might be physiologically relevant.
To answer your points directly:
manual observation on a per-micrograph basis, or large, large increases in box size (and losing lots of boxes as they clip off the edge of the micrographs)
Hi @rbs_sci thanks for the feedback! I have a high res map of the single particles! It appears that the stacked particles are physiologically relevant…
I would love to get at least a low-res reconstruction of 2/3 stacked particles, but I am running into issues during ab-initio classification. If you are interested see my other post: Ab initio reconstruction keeps failing with multiple errors
