Thanks for attending the webinar or viewing the video of the Open for (neuro)science tutorial: MICrONS Explorer with @sharmishtaa, @ceesem, and Leila Elabbady! This webinar demonstrated how to access and analyze the MICrONS Explorer electron microscopy data. You can view the demonstration Python notebooks used in this tutorial here.
Please feel free to chime in here with additional questions or follow-up questions to those asked at the live webinar. You can find the recording of this webinar and information about other webinars in this series here.
Additional questions to those answered during the webinar Q&A:
Question: How can one access the vasculature information?
Sharmi: We don’t have vasculature information identified separately. However the segmentation does separate out objects and a large number of blood vessels should be segmented out. But some of them will have errors by being wrongly merged with another object. You can click around in the 2Dview on a vessel to look at them. Creating a detector and/or database of vasculature is definitely a possible project!
Question: Can you color multiple segments simultaneously to the same colour in layer 2?
Casey: Yes, if have right clicked the segmentation layer and are looking at the id list, clicking the color rectangle box to the right of the id will bring up a color picker and you can actively choose the color.
Question: May I ask how you generated the list of soma positions? Was it manual?
Sharmi: The soma valence table that is downloaded with this tutorial were all manually annotated. However, we have an automated nucleus segmentation which we plan to release soon.
Question: is it possible to export information of the cell’s morphology and synapse location to use for computational modelling in simulators like NEURON?
Casey: The basic information is there for this in principle, yes, although we don’t have a simple exporter. There are skeleton and synapse files available on brain-map.org that would let one produce an appropriate file for NEURON. The one caveat is that in in the current volume, most of the cells have dendrites that have been cut off by the edge of the volume. However, we are currently working on datasets that have many cells with complete or virtually complete dendrites. We haven’t gone down this route ourselves yet, but if you are interested please give it a shot and get in touch if you have any questions!
SWC files can be found from the BIL Data Directory : Ultrastructural Connectomics - brain-map.org
Synapse data and the mapping between synapses and skeleton or mesh vertices is contained in the “meshwork” files which can be read by MeshParty.
Question: Is the processing in the cloud or locally?
Casey: Neuroglancer is a typescript app that runs in the browser. Each layer knows which cloud storage locations to go to grab the appropriate data for a given location in space or segment root ID.
Question: What is the overall purpose and goals of using state builders?
Casey: StateBuilders help you map dataframes to neuroglancer states, as well as simplify all of the details of generating neuroglancer states. By separating the rules of the mapping from the states, you can create generic rules and reuse them for many datapoints.
There are many options that I didn’t describe just now. Some examples can be found on its github repo here: NeuroglancerAnnotationUI/statebuilder_examples.ipynb at master · seung-lab/NeuroglancerAnnotationUI · GitHub