Tools to produce, analyse and compare both simulated and recorded neuronal datasets
I think getting the time to jump in a hangout will help us all to answer exactly that question
“Where are we right now?"
this looks exactly right in terms of the amount of neurons they removed
Check out this book: https://drive.google.com/open?id=0B_t3mQaA-HaMdFdGeFcxNEstZUU in particular the introduction, chapter 18 and chapter 10
Hi @/all — I thought this was interesting: https://docs.google.com/drawings/d/1x-aLpq9pnZ6ZuNznm426POV1cZwldQK7xYI9nOECAh0/edit?usp=sharing
@raminmh @lungd In the spirit of making iPython notebooks, I wonder if you guys can try to run this: https://github.com/openworm/simple-C-elegans/blob/master/README.md#quickstart — it should give you an environment you can use to both run and author notebooks and will load up a relevant example on your system
@raminmh @lungd this: http://hollopeterlab.vet.cornell.edu/uploads/7/3/7/0/73705987/optogenetics_in_worms.pdf is the paper I was referring to today that looks at muscle activation through synapse
I just came across this dissertation about probabilistic methods of analyzing neural spike trains with bayesian inference: http://www.columbia.edu/~swl2133/#thesis.
How have you been @theideasmith ?
Hi Dr. Larson. Thank's so much for your message. I'm doing well and have finally got back into the swing of things. The first semester I focused on school but now I'm getting back. Incidentally I'm going to be in the lab of Liam Peninski working on state space models for none other than Andre Brown's eigenworms.
These are some PCA space plots for the data used in the paper http://journals.plos.org/ploscompbiol/article?id=10.1371/journal.pcbi.1000028 Dimensionality and Dynamics in the Behavior of C Elegans
The burning question is how to connect them with these plots
@slarson how are you doing? I'm wondering if there are edge weights of the worm connectome. If not, perhaps they can be computed as the covariance of the mutual information between two neurons or something of the like. Why am I asking this? My idea is to find an embedding of the worm's brain in R^3 such that the distance between each node in space is the same as the weight of the edge connecting those two neurons. I then want to fit a surface plot to this embedding and interpret the flow of firing frequencies of neurons through the network as actually being particles moving through a space lattice approximated by the three dimensional embedding which is reconstructed continuously with computational procedures. If I can recreate the surface then one can define such things as "neural field equations" similar to electrodynamical field equations by considering the "potential energy" of a particle at each point on the neuronal space manifold.
Then, you can consider the physical properties of this neuronal space manifold such as
- How the neuronal space manifold changes over time
- The thermodynamical properties of neuronal space particles
- etc
@theideasmith I’m just seeing this! Holy cow this is cool. Please post over in Slack… we’re not on Gitter much any more! (Public link for folks is here: http://openworm.slackarchive.io/)