• The UI design of this project is based on Tkinter / ttkbootstrap
• The core module of this project is based on brainpy

• Make sure you have the correct python environment
• Execute instructions

Pyinstaller -F pysim.py 				# or
Pyinstaller -F -w pysim.py 				# Packaging without console
Pyinstaller -F -w -i XXX.ico pysim.py 	# Pack and specify exe icon packaging

run pysim.exe:

• On the initial screen, enter the MEA size (default 8 x 8)

• We can get the following main interface:

• Digital Twin Model:
  • Clik Digital Twin Model and choose data/2.1.1/2.1.1.mat
  • Clik Preset and choose data/2.1.1/args.yaml
  • Click the Run button at the top of the interface
  • We can get the results in logs/outputs:

The simulator page mainly includes 6 components:

• Mode selection: including Digital Twin Model and Virtual Experiment

  • Digital Twin Model: Fitting digital twin models with real biological data

    Preset: load existing parameters (.yaml file)

  • Virtual Experiment: Use existing digital twin models to conduct different virtual experiments

    Comp: Load real data to compare with virtual experiments (.mat file)

• Build stimulation paradigm: control which electrode emit stimulation

  load: Loading an existing MEA stimulation format (.txt file) clear: Clear the existing stimulation paradigm

• Core parameters (must be selected): select neuron type, synapse type, long-term plasticity model

• Run parameters

• Topology

• Other parameters

• Limited by brainpy and jax, only cpu can be used; Because of the use of cpu, the speed is slow and takes about 1~2 minutes;

• Due to jax, the packaging fails. Here is a simple code comment of the problem jax-ml/jax#17705 File "jax\_src\interpreters\mlir.py", line 711, in lower_jaxpr_to_module jaxlib.mlir._mlir_libs._site_initialize.<locals>.MLIRError: Verification failed

• As the number of neurons and synapses increases, more randomness is created; excellent results require more attempts under the same parameters

• Each data directory contains three files:

  • .mat raw pulse data file
  • .txt stimulation paradigm file, 8x8; can be loaded with the UI interface load tool
  • .yaml parameter configuration file, containing a set of fine-tuned parameters; can be loaded with the UI interface Preset tool
  • .pkl topology organization file, mainly used for simulation
  • outputs stores related simulation results