Pyr Science Questions

Pyr
#21

ok, thnx

2 Likes
#22

Three questions:

  1. Which dataset is currently more important to do? BANC or CA3?
  2. Is it ok, that we leave the end-of-branch annotations in the “Done” link, or would it be better, if we remove them?
  3. Are mitochondria necessary to add to the trace, or can we leave them, as in BANC?
2 Likes
#23

why does the AI exclude them when they’re inside the cell/neuron membrane is what I’m curious about. Going -strictly- by EW AI, she’d have included everything within that cell membrane whether it was mitochondria or a 200000 CB merger, lol.

1 Like
#25
  1. For now there is no explicit priority so feel free to do either, or if you feel so inclined you could work on whichever one seems like it needs more attention :slightly_smiling_face:
  2. Annotations in the completed link are fine, and potentially helpful
  3. I’m not 100% on this but I’m pretty sure it’s fine to leave them out. If you have a specific example that might be good for a more definitive answer here
1 Like
#26
  1. No need to add mitochondria to the trace. Please leave them like the BANC.

I believe the are segmented as the intercellular structures can be useful. It’s also related to how the ground truth for understanding the boundary lines is trained. So a thick black boundary line is separated out. It’s seen as better to have the mitochondria be their own segments than causing issues with other boundaries being missed.

3 Likes
#27

New structural question! What is this thing and its purpose? What kind of neuron does it belong to?

image
image1500×742 217 KB

NGL link to segment

3 Likes
#28

Oop, another “What’s going on here?”

https://play.pyr.ai/#!middleauth+https://global.daf-apis.com/nglstate/api/v1/6637763077079040

The 3d is a bit blobby, but essentially there’s two little nubs here that start as a single, split into a double, and then just end. Kinda like a compressed u shape.
image

3 Likes
#29

If I accidentally trace a segment, that didn’t belong to the starter segment/merger and leave it in the final link, would that be ok, or the links should contain strictly only the starter segments?

2 Likes
#30

What is this type of a cell?
Is this even a neuron or some type of a glia?

image
image773×647 129 KB

https://play.pyr.ai/#!middleauth+https://global.daf-apis.com/nglstate/api/v1/4600099310141440

3 Likes
#31

its something ive got to find and trace myself it’s what it is :joy:. Looks pretty.

1 Like
#32

cant be glia imo, no ‘star anise’ shape(s).

1 Like
#33

After some middle-of-the-night search I think, it might be a protoplasmic astrocyte, so, a glia, after all.
You can take mine, if you want :smiley:

2 Likes
#34

interesting, it doesnt have any of the shapes/2D+3D im accustomed to seeing in a glia. lol

and thanks!

1 Like
#35

It also doesn’t have any synapses, AFAICT.

1 Like
#36

AFAICT?

(I wish there wasn’t a character min. limit, or i’d not have written this as AFAICT? was what I had in mind. lol)

1 Like
#37

AFAICT = As Far As I Can Tell (lol)

1 Like
#38

ah, lol. That makes sense.

1 Like
#39

https://play.pyr.ai/#!middleauth+https://global.daf-apis.com/nglstate/api/v1/4880602164625408

image
image1078×578 85.7 KB

Slightly related to the myelinated axons, there seem to be a few cells that follow (eg the annotated (wip) in link above) or wrap/spiral around (eg the current position in link above) the myelin sheaths, although the segmentation for them is quite limited, which makes them take a while to trace out. They usually seem to terminate by surrounding the axon once the myelin section ends.
(Edit: went back to a previous link of mine to find a nicer example of the spiralling)
https://play.pyr.ai/#!middleauth+https://global.daf-apis.com/nglstate/api/v1/5675220019970048

image
image500×510 8.09 KB

This one has a bit more of its segments intact.
I guess i have 2 questions:
anyone know what these cells are? its interesting that they follow along so tightly in the way they do.
and how much work should we put into proofreading these for the ones with poorer segmentation? (maybe we could separate them out into their own task list when we come across the larger ones?)

5 Likes
#40

oh so pretty, helix neuron! lol. I’m almost certain that that’s a glia cell, one of those that wraps around other neurons to ‘transport’ nutrients etc.

If I am correct and it is glia you dont have to proofread it at all, but let’s see what an admin says abt this.

3 Likes
#41

yeah, i thought it was glia at first and did them for fun anyway, but seeing the “full” cell now in my current batch has me less sure :thinking: It’s very normal looking up until it starts its axon following.

2 Likes