Unfortunately, no full chandelier neurons in the CA3 region of the hippocampus… but they do innervate (connect) with the axons of Pyramidal cells. So it’s possible that we may proofread some of their branches that reach into this CA3 dataset 
2 Likes
How careful are we supposed to proofread the neurons, should we proofread every dendrite or just backbone. Ofc most of the branches are just straight from what we have so far, but there is a also some whole cells in the bunch.
1 Like
something little than 24hrs into this new-est game/dataset (Pyr) and one thing is certain: Pyr AI was much too jealous of MSTY’s crayons and got a titch tad too creative in colouring…all the things together. lol
3 Likes
Yeah it got a especially overzealous near the edges of the dataset 
3 Likes
most astoundingly so, yes. lol
2 Likes
Good question! Yes, please proofread the dendrites. You’ll find that in this dataset there aren’t cells with tiny little “twigs” for the dendrites. Please ignore the proofreading the spines except for the following rule that applies to removing small segments from the dendrite branches: You may ignore small dendritic mergers that are ≤ 5µm and have NO boutons. BUT we do ask that you remove any dendritic mergers with boutons.
To know how big a segment is in the 3D, hit the “O” key on your keyboard to show the scale bar in the 3D. The scale bar will resize as you zoom in/out.
Here’s a link with two old example cells that have some dendritic mergers on them that needed to be removed because they were bigger than 5µm: neuroglancer
I hope this makes sense!
3 Likes
I’d like to say, that it feels so good to see so many pyromaniacs (lol) working on this new dataset. Feels like the first days of Flywire, when we were exploring the new dataset, new working environment, scientific papers, etc.
4 Likes
Yeah! I had to drop off the fly brain pretty early on because I was starting a new job and other things took priority, hopefully nothing’s getting in the way this time around.
3 Likes
i think this beeing a smaller dataset, and a lot of straight branches, makes this dataset easier to work on so far compared to flywire. I also like the way we work with a assigned starting segment. but would really like some easier way to ask for help on segments.
3 Likes
this caught on fast! lol and yeah it’s nice to have the old team back 
H5 to every1!
3 Likes
I liked the bigger variety BANC had, having the entire spine and brain to work in, sooooo many different neuron types. But this dataset is defo easier to work with (other than everything being merged together lol) lots of basically straight lines lol
4 Likes