Scrambled mapping between bodyID and label for hemibrain

[Peter Pires]

I’m not sure if this is the appropriate venue to raise this issue, but using the live demo to view the janelia hemibrain data, the mapping between bodyID and label is scrambled (but the mapping between bodyID and segment is correct). For example, Below link should show EPG neurons, but instead shows  random assortment of bodyIDs:
https://tinyurl.com/xthjzjs9

[Viren Jain]

+Berg, Stuart 

On Mon, May 24, 2021 at 9:18 AM Peter Pires <p.musse...@gmail.com> wrote:

I’m not sure if this is the appropriate venue to raise this issue, but using the live demo to view the janelia hemibrain data, the mapping between bodyID and label is scrambled (but the mapping between bodyID and segment is correct). For example, Below link should show EPG neurons, but instead shows  random assortment of bodyIDs:
https://tinyurl.com/xthjzjs9

--
You received this message because you are subscribed to the Google Groups "Neuroglancer" group.
To unsubscribe from this group and stop receiving emails from it, send an email to neuroglancer...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/neuroglancer/d6c4ff9f-bfd9-4ae9-b1ac-79e3ea23d888n%40googlegroups.com.

[Jeremy Maitin-Shepard]

How did you obtain that particular view?

Note that in the segment list on the right, the query is set to "EPG", which is why the list shows EPG neurons.  However, you will note that those checkboxes are not selected --- because none of the neurons in that list are shown.  To see the neurons that are shown, delete the query, then scroll to the top of the list.  You will see that indeed a random selection of labels is shown.

To actually display the EPG neurons, first deselect the existing neurons by unchecking the checkbox labeled "50 visible segments".  Then click the checkbox where it says 50/22706 matches to select the EPG neurons.

To view this discussion on the web visit https://groups.google.com/d/msgid/neuroglancer/CACveBs4Xpx7-YKXxV_YUvTVPBFcVwJz9e6UScrp7maJbaRbzyQ%40mail.gmail.com.

[Peter Pires]

I pasted https://neuroglancer-demo.appspot.com/ into my browser, then added precomputed://gs://neuroglancer-janelia-flyem-hemibrain/v1.2/segmentation as a segmentation layer. Then, I queried for “EPG” and selected all 50 matches, this produced the following URL:

https://neuroglancer-demo.appspot.com/#!%7B%22dimensions%22:%7B%22x%22:%5B8e-9%2C%22m%22%5D%2C%22y%22:%5B8e-9%2C%22m%22%5D%2C%22z%22:%5B8e-9%2C%22m%22%5D%7D%2C%22position%22:%5B17216.5%2C19776.5%2C20704.5%5D%2C%22crossSectionScale%22:1%2C%22projectionScale%22:65536%2C%22layers%22:%5B%7B%22type%22:%22segmentation%22%2C%22source%22:%22precomputed://gs://neuroglancer-janelia-flyem-hemibrain/v1.2/segmentation%22%2C%22tab%22:%22segments%22%2C%22segments%22:%5B%221011335902%22%2C%221037895459%22%2C%221069266255%22%2C%221131831990%22%2C%221141682748%22%2C%221163185756%22%2C%221164209179%22%2C%221165085220%22%2C%221230738118%22%2C%221231619255%22%2C%221232607453%22%2C%221252337419%22%2C%221279775082%22%2C%221281618303%22%2C%221313853933%22%2C%221386780576%22%2C%221387583729%22%2C%221387881882%22%2C%221474757586%22%2C%221477447226%22%2C%221477774797%22%2C%221478400968%22%2C%221478487101%22%2C%221543717447%22%2C%221670256170%22%2C%221697608860%22%2C%221697617556%22%2C%221697621973%22%2C%221697949646%22%2C%221698338958%22%2C%221722333142%22%2C%221753359080%22%2C%221783362106%22%2C%221784057323%22%2C%221809264255%22%2C%221934210967%22%2C%221950304963%22%2C%221950641537%22%2C%222010246349%22%2C%222011364502%22%2C%225812990925%22%2C%22663298610%22%2C%22675800901%22%2C%22706520453%22%2C%22735112319%22%2C%22796052404%22%2C%22796755936%22%2C%22847297982%22%2C%22974455281%22%2C%22977667682%22%5D%2C%22segmentQuery%22:%22EPG%22%2C%22name%22:%22segmentation%22%7D%5D%2C%22selectedLayer%22:%7B%22visible%22:true%2C%22layer%22:%22segmentation%22%7D%2C%22layout%22:%224panel%22%7D

The segments that are displayed are not EPGs, and if I hover over a neuron it will display “EPG” even if it isn’t. It does look like the list of segment bodyIDs is correct though…

[Jeremy Maitin-Shepard]

Hi Peter,

Thanks for reporting this.  Turns out this was a bug that I introduced recently in Neuroglancer, as part of some improvements to the segment list display, not a problem with the data itself.  I have now deployed a fix, though it may take a few minutes for it to take effect.

To view this discussion on the web visit https://groups.google.com/d/msgid/neuroglancer/f8e0f4f9-bb59-4b40-a07c-49b807833a40n%40googlegroups.com.