Proofreading in the BANC

Welcome Flyers to the BANC!

Quick list of the resources provided in the Access Granted email:

Spelunker Specific Shortcuts

P.S. Want to practice the new edit commands in Spelunker before editing in the BANC? You can use the FANC (nerve cord only) practice sandbox. The EM will look a little bit different from the BANC, but you can practice with the new edit commands worry-free! :slight_smile:

Any other resources, quick references or links you’d like us to add? Let us know!
Please feel free to ask any questions about proofreading, how to declare a neuron is “complete” using the banc-bot, a weird neuron structure or anything else!
For Science!


Just a reminder that we’ll be hosting an Intro the the BANC Zoom session next Thursday, April 25th at 10 am US ET

Let us know if you’d like to join and we’ll send Zoom info. Your video/mic does not need to be on to participate.


Just wanted to add to the shortcuts, that while double clicking a segment (either in 2D or 3D) only hides it, you can press Shift+double click to remove the segment completely.


Same day as Fallout 4 update :stuck_out_tongue:
I’m interested in receiving a link for the Thursday BANC Zoom session :+1:

I’m not already into it for the moment, i’ve to understand how it works as i have not been long enough on Flywire.
Is there a short video to show how to proofread exactly ? I mean, like someone screening his own session with “loading - reaping - proffreading - sending - next” ?
It would be helpfull for me and other people i think !

Thaaaaaank you !

PS : I’m totally sorry, my browser didn’t load the links in first post. i’ll get an eye on links !

1 Like

Is there a way to increase mesh detail ?

Also, do we have to remove small segments (like eyewire dusty nubs) or these are not relevant ?
Like this in green :

Or only real nubs and branch are relevant ?

I think, there isn’t any way to do it. If you go to the Render tab in the right panel, look at “Resolution (mesh)”. As you can see, there’s only one vertical bar showing available resolutions (there’s also the gray bar, but it just shows your current settings - whenever you place it, the closest value will still be that one red bar, so it doesn’t matter, where you place it).

As for the small segments - I believe, those don’t matter. Also, you don’t have to add small segments. Take a look at screenshots in Expert Proofreading Tips for the BANC (Brain and Nerve Cord) Dataset - FlyWire Blog in the section titled " Knowing Your Cell Is Complete". There’s also some basic info, how much you should care about the details.
Also, take a look at Jasper’s video showing, how to work with the BANC dataset. As you can see, he only looks for major missing parts and clearly visible mergers:

1 Like

Nice and clear answers, i always like that with you KK ! :ok_hand:
So i focus on missing branches and eventual “interesting” nubs.
Next step is to finish my first cell and complete it. As i understand it’s not the same way than eyewire.
I’ll take a look on your link, thanks you :hugs:

1 Like

You’re welcome!

To complete a cell, copy its coords (right at the top, the double square icon). Go to Slack, switch to banc-bot, and post something like this

134330, 50991, 2366! backbone proofread

the coords, obviously, should be yours. Don’t forget the exclamation mark.

1 Like

I finished my first one after understanding all the reap workflow. I’m on my way to complete but Is it normal this green neuron is not going inside the red rectangle ? Or rectangle isn’t important ? As it end as a big nub, i didn’t find any extension. I’m not used to this kind of neuron.

After receiving Jasper autorisation, my first cell is done !
I seen there’s a lot of possible annotations to add, like “! descending” “! ascending” “! spans neck”
Is that a thing i should also do or citizenonly need to “complete” ?

Hi Kfay,

Would you mind sharing a link to your cell? (You can click the “Share” button to autocopy the link in the top right bar same as the “Share” button in FlyWire).

Is there a soma/cell body? Neurons that run up and down the neck connective may have more arbors/branches at the ends. You can see a sample selection of completed Ascending Neurons and Descending Neurons in the BANC that travel up and down the neck here.

Re: the red rectangle – We’re asking folks to begin proofreading on the left side of the GNG (Gnathal Ganglia), as indicated by the red bounding box in the starter link. You can click and select any neurons to start proofreading in that zone, but it’s okay if you’re finding other cells to proofread in the dataset!

PS. You can find some GNG resources and overview in the Intro to the BANC for Citizen Scientists post.


Currently, just use the “backbone proofread” tag to complete cells.

If you’re fairly certain you know a cell type, then you can add the name/annotation. (I think for most Flyers the cells in the Optic Lobes are where folks are most comfortable adding cell types). If you’re unsure, please do not worry about adding the other annotation tags right now.

1 Like

Finished my 1st two cells as well! I’m liking the new dataset, whole new stuff to explore lol.

Imma farm them up like I did in FW. lol

1 Like

Well i gave up on this cell, letting it for other users. I prefer being totally confident on cells i’m on for the moment. So, sorry i do not have the ID anymore.
I already complete 3, but i took them from banc-bot with “todo” (like Jasper demonstrate how to do in his youtube video). I’ll give a try on GNG as you mention !

For grabing a neuron on the Leftside GNG, i just have to right clic somewhere on the box for pointing and double clic on any part in 2D image to select a neuron, right ?

Ok for the proofread, i’ll stay on “backbone proofread” tag for the moment and learn more about conditions. I’m really enjoying these stuffs and want to learn more !

Sorry for all these questions, but i’m dyslexic and with all these informations forum/blog/youtube etc it’s a bit too much to understand correctly.

An thank you again for your time !

1 Like

I’m completing my first GNG cell, could you please verify it and tell me if the intergrity is good ? There’s no Cell Body and i’m not used to this kind of cell.!middleauth+

coords in case needed : 108092, 59363, 2195

Didn’t merge it, but found the CB!middleauth+

link’s 2D should be right at the connection point. Haven’t done many of this cell type yet either but the few I’ve done all had the CB connection point doubling back near the dendrite end of the backbone like this one.


Pretty nice, thank you ! That’s indeed pretty tricky, wasn’t at all looking for this kind of connection !
That’s a new arrow to my bow :+1:

1 Like

yw and yeah idk abt this dataset but the previous FW dataset was full of tricky and/or basically nigh-non-visible-thin connections like some of the BP ones in EW (the kind atani used to be good at finding lol). The fact that we can’t change axis here like in ew doesnt help.

1 Like

Yeah, it’s kinda very thin and eye broker.
How do you guys do without possibility to highlight parts already checked ? my actual cell is so big, i’m loosing myself in “what did i do” … :melting_face:
Is that allowed to split sections for a convenient workflow ?