i have submitted a few where i belive it to be Tm 8, but not 100% sure and have written it down as Tm looks like Tm8 but not sure if that is smart or not, it is weird that they do not have a better way of naming the cells that could be sorted easier
For instance first column general type (tm, t, Dm, unidentified etc) next column variant (1, 2,3a, unidentified etc) + a box for how sure you are about each case + a box for free text, preferably the types should be from a list to make sure spelling was good.
if you could add type is Tm without being sure of the variant some one could then afterward look at all tm cells and compare them to each other, but as it is now i do not really see how they can make a query to find what is really completed or not.
Thanks for the info! Idk maybe they have admin tools/things we cant see that allows them to? in any case ive not yet tried to actually to submit a cell identification, i’ll find a cell or two im 100% sure of type and variant to submit.
Listing the cell class (for example TM) is more useful than no identifying information, so you can still submit even if you don’t know the exact type (TM1, TM5 etc). We also ask that you type in the full name in addition to the abbreviation.
You’ve got it right that we can later go cell class by cell class and add cell type information. I was just in meeting and a stat from central brain that was shared mentioned 18K neurons with class label but only 5K with type. So that’s essentially already happening.
How important is it to add the full name? considering how much difficulty i have had with typing in the names in the Q&A sheet, using time on this for every cell is not something i am willing to do. I understand why it would be needed and see several reasons for it, but would rather use my time on the tracing.
Maybe it would be possible for devs to add a list of all the optic lobe cell types to the identification dialog and we would choose a type from the list and only type a number or number and letter (e.g. 1, 15, 3Y).
Yeah that’s a good idea. The only challenge is that the cell ID popup is global so all the users see it and most are working in central brain vs optic lobe. I wonder if it might be possible to do it through a script?
By the full name do you mean typing out the full name vs the shorthand, or do you mean labeling the cell category type vs the specific cell?
For example:
“Medulla intrinsic” vs “Mi”
or
“Medulla intrinsic 1/Mi1” vs “Medulla intrinsic/Mi”
@annkri T2a was missing from the spreadsheet. I’ve added that and 2 potential examples of such cells, but I’m not 100% sure, if they are correct.
I’ve also added a link for TmY5. There’s an entry for that in the Cell Completion tool, but its description says “Not enough time to proof”, so I’ve proposed mine, just in case.
If you or anyone else could double check those links, that’d be great.
Question: We (players) have encountered (so far) a few Somas w/o a backbone (cant be found) and/or Backbones w/o a soma, would the researchers prefer those cells to be left as they are uncompleted and un-classified or would it serve to complete them to get them out of circulation and give them a ‘unknown soma/backbone missing’ classification so that later on either we (players) or researchers can search the DB find those cells w/o soma/backbone and attempt to deductively find soma/backbone by what’s left (much later when we’ll have completed almost 100% of lobe(s)) untraced?
But I didn’t found anything during checking it.