Someone on the c4d science animation google group had recently asked about astrocyte data. If I remember correctly there might have been a solution. I build those and microglia by hand which as a result end up slightly stylised. It would be cool to have the actual data files.
NeuronBuild 1.8
ScientiaViz
#8
Thanks Joel.
I tried to get into that Google group but couldn’t get in for some reason.
Something for glial cells could come in handy, though. I built this astrocyte by hand from ortho images. Fun but not that fun. Journal of Neuron cover.
NWoolridge
#10
Great image, Arik!
Here are some of the resources I was thinking of. These were passed on to me by Amy Robinson of Eyewire.
Neuroglancer is a project by Google to provide an interface for volumetric visualization in the browser. There are some very cool demo datasets on the github page worth exploring.
Museum with retinal cell meshes: http://museum.eyewire.org/?neurons=26065,20117,26051,17212
A number of great dense multi neuron datasets here in the gallery, and you can download the meshes…
ScientiaViz
#11
Hey thanks, guys.
Nick, very cool links! Definitely going to spend some time looking through these.
Incredible what can be done nowadays.
teknow
#12
Nick, Thanks for that link! Gorgeous
When I download the meshes I get only single neurons which seem no better than NeuronBuild.
How can I create actual multi-neuron meshes from the data sets? (besides making fake ones with cloning)
bullit2
#13
https://neuroml-db.org/gallery
The Neural Open Markup Language project, NeuroML, is an international, collaborative initiative to develop a language for describing and sharing complex, multiscale neuron and neuronal network models. The project focuses on the key objects that need to be exchanged among software applications used by computational neuroscientists.
Examples of these objects include descriptions of neuronal morphology, the dynamics of ion channels and synaptic mechanisms, and the connectivity patterns of networks of model neurons. This modular approach brings additional benefits: not only can entire models be published and exchanged in this format, but each individual object or component, such as a specific calcium channel or excitatory synapse, can be shared and re-implemented in a different model.
The NeuroML Database is a relational database that provides a means for exchanging these NeuroML model descriptions and their components. One of our goals is to contribute to an efficient tool chain for model development using NeuroML. This emphasis allows the database design and subsequent searching to take advantage of this specific format. In particular, the NeuroML database allows for efficient searches over the components of models and metadata that are associated with a hierarchical NeuroML model description.
But they have Blender models. So i think you have to use Blender to convert for C4D
http://blenderneuron.org/docs/introduction.html#overview
Edited: to get simulate voltage potential propagation in Blender instead of just mesh you have to install Neuron
https://neuron.yale.edu/neuron/ maybe someone can make an exporter for C4D
NWoolridge
#15
Thanks for the pointer to NeuroML; it would be great to be able to add simulated signal propagation and other aspects of neurophysiology to C4D…
teknow
#16
IMHO NeuronBuild is fun and NeuroML is enough to give anyone a migraine! If anyone thinks NeuroML is fun I would love tp hear about it and be converted.
demis
#17
Seeing your Vimeo tutorial my eye caught a plugin called autoPACK…
I’ve downloaded it but I really don’t understand anything … The site was claiming the plugin can pack anything not just proteins … Am I missing something ? Could you upload a simple scene <=R20 of objects packed together using this plugin so I can understand how it works ?
I was looking for a tool that clutters/stacks objects together. Can autoPACK do the job ?
teknow
#22
celke, These links are especially wonderful because Neural Science is one of my favorite areas to obsess.
Thank you!
NWoolridge
#23
Demis: Autopack was created by Graham Johnson as a means of packing arbitrary shapes and sizes of proteins into cellular compartments. In principal the algorithm should be able to pack just about anything, but I think that Autopack is pretty much set up to package “recipes” of cellular molecules.
You could get a similar (but not nearly as efficient or correct) effect by using rigid body dynamics with a shrinking, jiggling container with a lot of objects in it…
NWoolridge
#24
Teknow: Tyler Sloan has created a neuromorpho importer for Blender:
Haven’t tried it myself, but I plan to…
demis
#26
I was hopping for a less simulational (non time-dependant) algorithm 'cause rigid body dynamics can get heavy if the objects are many or/and have many polygons.