Done.
(more characters b/c it doesn’t like only less than 20 ignore this parathesis lol)
3 Likes
@admins what do we do if we find already identified cells are wrong?
i would say that all of the ones here Sign in - Google Accounts
is lawf2 cells since fishback is not showing that second arbour going upward on any of the mi cells, not sure about first entry by lucy since that one is just ending in a nub
4 Likes
i find that if i add more cells i am unsure of too the same tab with for instance Tm cells it would be easier to compare different cells and will probably turn in more of the same type. when we then find out what type of cell it is we could identify all of the ones with the same type instead of having a separate entry for each cell. but it ofc make it difficult to have a discussion on the gsheet.
2 Likes
i wonder if we couldnt do this, not the mis-identification bit lol, but in that link in your post the first cell is “Mi editing not complete” so i wonder if we couldnt do the same with some of our cells, ie: C2 Soma missing" etc.
You are thinking about the ones in the black spill area? think the admins did already say we should not complete those. the ones i have seen with comments about missing soma, not completly proofreed etc looks complete, but might have some corrupted volume where part of the soma is or something like that. The ones that are not completly proofreed looks complete to me.
2 Likes
yeah those and yeah i know, but hey if the researchers are doing so…? I mean i wont w/o permission but was just wondering/thinking about it.
Random aside here, I fixed the setting so that the minimum is 7 characters for a post now 
6 Likes
lol, than you.
1 Like
Medulla Intrinsic Mi8
I spent a while this morning on examples of the Mi8 cell, and I thought I’d share my experience here, to hopefully save others some time. It looks like the representation of Mi8 in the Fischbach paper (or at least, my interpretation of it), is not representative of what we actually find in our dataset…specifically the axonal arbor, which I spent quite a bit of time on, this morning. The picture and link below contain what looks to me to be an actual Mi8 cell (yellow), together with its associated Mi1 (red), Mi4 (green), and Mi9 (blue). The diagram in the Fischbach paper suggested to me a “bleb” type axonal arborization, but what I actually found in the dataset looks quite different, to me. So…just thought I’d mention it, and save others some time.
Also, I’d just like to point out that even though we have a gsheet devoted to unidentified cells, it is still okay to discuss cell type identification here (kinda seems like the gsheet killed the conversation here, lol). The gsheet is actually an excellent supplement to this thread and conversation, to prevent questions asked about identifying a particular cell from getting buried in all of the discussion here. So…just pointing that out, fwiw.
Cheers, all.
https://ngl.flywire.ai/?json_url=https://globalv1.flywire-daf.com/nglstate/5115769021005824
4 Likes
Yeah, I’d like it to be:
ghseet - concrete completed cells, which people have problems with identyfing,
"Visual Cell Type Illustrations" thread - mainly pictures and drawings of various cells. Especially those types, that aren’t in Fischbach and Dittrich, so good for quick glance over various types,
this thread - more text descriptions and informations about various types. More in depth discussions,
"Useful Structural Patterns in the Optic Lobe (inferring microstructure from macrostructure)" - descriptions of structures that contain more than a single neuron.
My cell identyfing process is:
- Fischbach and Dittrich (the FW version)
- first post from “Visual Cell Type Illustrations”
- the @annkri’s gsheet
- the rest of “Visual Cell Type Illustrations”
- this thread + “Useful Structural Patterns in the Optic Lobe (inferring microstructure from macrostructure)”.
Also, these are probably actions from fastest to slowest (takes more time) to digest.
5 Likes
@annkri That example of T2 in gsheet looks to me more like T2a. T2 usually has the axonal end wider and also has that characteristic small middle bushy segment in the dendritic neurite (look at the examples in the link in H column and Fischbach’s paper). Not 100% sure though, so I won’t change it.
1 Like
@kk i agree with you, that T2 cell is from the guide we are supposed to use, but can ofc be wrong there also
3 Likes
changed the sheet, probably more mistakes in there, when looking closly
2 Likes
Cool. And we’ll be looking and fixing them one by one 
2 Likes
Should we create/specify a type Tm2a variant, in the examples gsheet and in the identification database?
So…I came across this Tm2 cell, whose arbor in the proximal medulla is separated from the main backbone by a “stalk” (as Fischbach calls it), in exactly the same way as the Tm1a variant of Tm1 and Tm18a variant of Tm18 (reference the text in Fig. 8 of the Fischbach paper). Although there is no Tm2a variant documented in the Fischbach paper, this seems to be a similar variant of Tm2.
Any thoughts on whether it would be beneficial to distinguish this morphology as a separate subtype?
https://ngl.flywire.ai/?json_url=https://globalv1.flywire-daf.com/nglstate/6487263747768320
2 Likes
okay i added some new rows for the Tm1a, Tm2a, Tm4a, Tm18a
and added in pictures from your cell. i know that i have seen some examples on Tm1a in one of the post, but do not remember where so can´t add it too the sheet.
I have too admit that i have a great deal of trouble to see the difference between the different Tm cells, and take a lot of time to find celltype and for now i just add them to the sheet since i think it will be easier to compare when seeing many cells at the same time. And plan to use annotation on cb to mark type when discovered.
2 Likes
@annkri Yeah, same here…it can be very challenging, in my experience, to sort out all of the many Tm and TmY cells. On my end, I have been slowly compiling my own collection of each of the different types, as I find them, and will begin linking my collections in your examples gsheet, once I have enough examples of each to provide a clear illustration of their respective morphologies. But yeah…so far, the Tm/TmY cells are the most challenging to identify, to me.
I have not personally come across an example of the Tm1a variant yet, in our dataset, but it is documented in Fischbach, so I expect that we will eventually find one/some/many.
It sounds like you are in favor of specifying a Tm2a variant, as a subtype. If no one has any objections, I was thinking that we could specify it in the database as, “Potential Tm2a variant of Tm2”, or something like that?
Thanks as always, Anne.
1 Like
I do like the idea of a subtype; however I would caution using a ‘standard’ type name for it as the cell naming conventions are established through published literature through the scientific community. In other words, if we call/label this cell type ‘Tm2a’ in our identification, it may/will conflict with later research and publications.
In cell identification, I would call this “Putative Tm2” and add a note that “the arbor in the proximal medulla is separated by a stalk”.
If anything perhaps HQ can okay us to have a Tm2 (FW)a type classification which indicates that the classification has come from citizen scientists.
I just want to make sure we are on board with the whole scientific community; even if we do discover and name cells in our work here, we need to be going through the appropriate scientific process before we give them a possibly confusing name/label.
1 Like