Are there any standard tools for neuroscience models programming, like for spiking neural networks?

Question

I come from deep learning environment and there is quite a few standard tools and frameworks that you can use in different occasions. I am trying to build a spiking neural network model in Python and I wonder if there are any standard frameworks and tools for visualization that I can leverage.

To give an analogy, there is TensorFlow or Pytorch in deep learning that facilitates building layers of neurons and run experiments. Is there any framework for spiking neural networks? Tools to automate calculatinos per neuron or run iterations?

For visualization I need also another tool that visualizes the spiking neural network and its evolution during learning. Is there any tool that is tailored for visualization of such these networks?

Many thanks!

Answer 1

Brian2 is a great tool in python for spiking neural networks: https://brian2.readthedocs.io/en/stable/

NEURON might feel a bit old and clunky but if you want more biophysical detail, it is great: https://nrn.readthedocs.io/en/8.2.4/guide/index.html. It is mainly written in C++ but can be called from python.

Tangential but might be useful: For general differential equation solving using numerical methods, I am in love with Julia's DifferentialEquations package: https://docs.sciml.ai/DiffEqDocs/stable/.

Answer 2

A way into your problem might be to look at the NeuronDB website, which shows models for cell types, genes, transmitters, simulators, and many other search variables. You can then narrow down your search for a Python model.

https://senselab.med.yale.edu/ModelDB/FindByRegionList