I4 Symmetry Missing from symmetry.py File

Troubleshooting
1

Hi,

So this is an easy problem, but I’m not sure about the fix? I kept getting the error below:

AssertionError: Missing symmetry file or unrecognized symmetry type: I4

while refining the ab-initio model of an icosahedral particle. Looking into the code with nano, it turns out that my symmetry.py file has I, I1, and I2 symmetry, but is completely missing I4.

I’ve tried updating the node/files to no avail. Is this a known issue, is there a standard fix, or can someone just post the required code? (I2 code included below as an example)

‘I2’:
“”"
rot_axis 2 0 0 1
rot_axis 5 -1.618033989 -1 0
rot_axis 3 -0.53934467 -1.4120227 0
“”",

2

I got the error “Missing symmetry file or unrecognized symmetry type: I3” during my Ab-initio job. It’s odd that “L3” worked.

3

Hi @tosyl!

I have two questions, if you don’t mind!

First, could you explain a bit why you need I3 symmetry as opposed to using one of the other icosahedral symmetry conventions (I/I1 or I2)?

Second, I have a question about

Was that a typo, or did entering “L3” as a symmetry operator work?

4

I read an article (Cryo-EM structure of severe fever with thrombocytopenia syndrome virus | Nature Communications) describing a single round of 3D classification using I3 symmetry. This process identified a major class with 90,188 particles and produced a 6.7 Å resolution reconstruction after 3D refinement.

When I attempted to replicate this process in CryoSPARC, using I3 symmetry (uppercase “I”), I encountered an error message: “Missing symmetry file or unrecognized symmetry type: I3.”

As an alternative, I tried L3 symmetry (uppercase “L”) during an ab initio job. While L3 worked, it was very time-consuming, so I terminated it.

5

I suspect that it wasn’t actually using a symmetry other than C1, but I’m surprised it didn’t throw an error when passed “L3” as a symmetry operator…

Also, @rposert, I3 used to be necessary for some things (the original implementation of block-based reconstruction was dependent on it, for example) and, if I’m honest, I like having the fivefold symmetry axis along Z. :wink:

6

Hi @tosyl and @rbs_sci!

Thanks for the additional info on I3 symmetry. We’ve recorded adding I3 and I4 as a feature request.

2 Likes