Skip to content

Latest commit

 

History

History
53 lines (24 loc) · 2.44 KB

README.md

File metadata and controls

53 lines (24 loc) · 2.44 KB
Raw

Analysis of Published Multicompartmental Models

Several Published emodels have been analyzed to see if they can be implemented to BBP pipeline https://modeldb.science/modellist/257

By checking this we know what are the necessary features for a successful emodel and also datasets used to tune. The github repo here tries to replicate the models and documents the results in notes.md

Alle et al 2009

Notes: This study models the axonal mechanism and how energy consumption of Mossy Fibers are not as high as previously thought by HH assumption.

useful for:

Mossy Fiber axon modeling and parameters for conductances , amplitudes etc Experimental traces for MF . Fig 1

ModelDB: https://modeldb.science/135838

Aradi & Holmes 1999

Notes: According to Beining et al , this was the gold-standard model before their paper. They indicate models dont behave correctly outside tuned region. However, Beining model is also not behaving consistent with different current protocols. When the modeldb was checked, it seems there are too few mod files and most of the channels are constructed with channel builder function of NEURON. So its not a good candidate unless those channels are exported as mod files.

ModelDB : https://modeldb.science/116740?tab=2

Beining et al 2017

Notes: Featuring T2N toolbox and using DGsg_GC as an examplar case. Its an improvement on Aradi & Holmes, 1999 . The detailed analyses can be found in examplar codes written in MATLAB. Since reviewers asked them to restructure the paper as a toolbox paper, they removed some extensive analysis and discussion on the GC model.

I cloned the model in a github repo, and reported my findings here.

ModelDB: https://modeldb.science/231818

Ferrante et al 2009

Notes: Analyzed FFI for IO response of DGsg_GC with MOPP cells. Used double exponential time courses AMPA, NMDA and GABAa synapses to buffer IO regime.

The code to replicate is in hoc format. I tried to convert it to python here. TODO: Needs adjustment.

ModelDB: https://modeldb.science/124291

Code Examples :

  1. example2.py : loads a morphology and analyzes dendrites over path distance. injects passive current mechanism and runs a current clamp to check signal propagation within neuron
  2. example3.py : defines a cell class and imports a given SWC morphology file , plots the neuron and injects a current after inserting hh mech to the soma and axon and passive mech to the dendrites