Skip to content

Home

Jump to bottom
William Silversmith edited this page May 9, 2017 · 11 revisions

Welcome to the neuroglancer wiki!

How to add a layer to neuroglancer.

Completing the Source

First choose the API that your data server satisfies:

  • brainmaps://
  • ndstore://
  • dvid://
  • render://
  • openconnectome://
  • precomputed:// (read more)
  • python://
  • nifti:// Then specify the url to your data server.

Examples

  • python://http://localhost:8888 A server running locally on port 8888 that follows the "python" API to serve data. Caveats: If you are accessing neuroglancer using and https connection, it will refuse to pull data from and http connection.
  • precomputed://https://storage.googleapis.com/neuroglancer/s1_v0/image A server that follows the "precomputed" API to serve data. This data resides in google cloud storage in as bucket call neuroglancer all the files in the image "folder" need to be publicly available.If you are feeling lazy at typing you can equivalently write precomputed://gs://neuroglancer/s1_v0/image
  • precomputed://https://storage.googleapis.com/neuroglancer/s1_v0/image A server that follows the "precomputed" API to serve data. This data resides in google cloud storage in as bucket call neuroglancer all the files in the image "folder" need to be publicly available.If you are feeling lazy at typing you can equivalently write precomputed://gs://neuroglancer/s1_v0/image. https is a requirement for http2 which enables faster download of data chunks by reusing the same connection.
  • precomputed://https://neuromancer-seung-import.appspot.com/blob/s1_v0/image Similarly as before, but if the data in the "folder" is not publicly available, our server at https://neuromancer-seung-import.appspot.com/blob is allowed to served it to authorized users. If you don't want to type this much, you can equivalently write precomputed://glance://s1_v0/image
  • AWS S3 example: precomputed://https://s3.amazonaws.com/neuroglancer/pinky40_v8/image

Completing the Name

Name doesn't matter, choose something you like.

Creating a Point or Synapse Layer

There's no graphical support yet for this, but if you're comfortable modifying the neuroglancer URL, you can do it!

Here's an example URL where we show the channel layer of the golden_v0 dataset. Most of the URLs you will see will have other information to indicate such things as the camera pose as well but we skipped that here for simplicity:

https://neuromancer-seung-import.appspot.com/#!{'layers':{'image':{'type':'image'_'source':'precomputed://gs://neuroglancer/golden_v0/image'}}

Now if you want to add a point layer, add the following layer JSON (with commas turned into underscores, ',' -> '_'):

_'NAMEOFLAYER':{'type':'synapse'_'points':[[$X,$Y,$Z]_[$X,$Y,$Z]]}

Note that NAMEOFLAYER can be whatever you want and you can specify any number of points. If you change the type to synapse you get a synapse layer. If you change it to point, you get a point layer. The major difference between them is that in a point layer, all points are the same color. In a synapse layer, every other point alternates color to indicate pre and post synaptic partners (and there's a line between them on some OS & browser combinations).

Here's an example of a finished URL:

https://neuromancer-seung-import.appspot.com/#!{'layers':{'image':{'type':'image'_'source':'precomputed://gs://neuroglancer/golden_v0/image'}_'points':{'type':'point'_'points':[[1020.5_1017_128]]}}_'navigation':{'pose':{'position':{'voxelSize':[4_4_40]_'voxelCoordinates':[1024_1024_128]}}_'zoomFactor':4}}

How to create self signed certificate for your local graph or state server.

openssl req -x509 -nodes -days 365 -newkey rsa:2048 -keyout privateKey.key -out certificate.crt

You need to tell your browser to trust your keys keys, go to an url served by your sever via https. Your browser will display a warning site, where you can click on advance -> proceed to insecure site. This way when neuroglancer uses your local server under the hood the requests won't failed.

Clone this wiki locally