[Chance-Device]

Incidentally, anyone know what is going on with this image - “Cryo-EM map of a center slice of the ushikuvirus particle”: https://journals.asm.org/cms/10.1128/jvi.01206-25/asset/1357...

It’s one quarter of an image flipped horizontally and then vertically, you can see the patterns.

It’s a bit odd to do that? Shouldn’t it just be the original EM image?

[observationist]

https://www.sciencedirect.com/science/article/pii/S104784772... - there are similar results in this paper, too.

After a bit of digging - it looks like it's done to sharpen features as one of the standard steps in producing these images. Where there are rotational symmetries in the things they're looking at, they focus on the smallest unit, and then rotate accordingly. If you had a trilateral symmetry, or hexagonal structure, they'd rotate 3 or 6 times around the center.

You're not getting a real image of the thing, but apparently it's got data from those other segments mixed in with the rotations, so you're getting a kind of idealized structure, to make the details being studied pop out, but if you have some sort of significant deviation, damage, or non symmetric feature it'll show up as well.

It's called "imposed symmetry" https://discuss.cryosparc.com/t/what-is-actually-occuring-wh...

Neat stuff, cool thing to catch!

[Terr_]

So kind of like taking a picture of a human, and then taking each half, flipping along the midline, and blending to get an idealized Symmetrical Human?

[RicoElectrico]

According to this article the image is computed and not really directly captured https://www.chemistryworld.com/news/explainer-what-is-cryo-e...

[Bjartr]

I think that might just be the original and it simply is symmetrical to that degree. I found a few more examples of "cryo-em center slices" and I've yet to find one that doesn't have really strong symmetry down to the small dot patterns.

A different paper, this figure shows a number of cryo-em images, including a simulation, and they all show the same degree of pattern symmetry https://www.researchgate.net/figure/Central-sections-through...

First figure in this third paper also shows symmetry of small patterns https://journals.asm.org/doi/10.1128/jvi.00990-22

[Chance-Device]

Thanks, those examples make it pretty clear.

I still think it’s super weird that it looks exactly like an EM image, but is generated. Anyway, good to know!

[jiggawatts]

Rampant fraud in science papers has reached the point where hobbyists can point out obviously fake charts and graphics even in prestigious journals.

Publish or perish needs to end.

[jibal]

This isn't fraud ... see the informative comments nearby.