A neural simulation app
As you know from personal experience, the brain is capable of a myriad of computations. You can control and coordinate fine moves to play the piano or carve your skateboard. You can feel a burst of joy when your team wins. You can learn neuroscience. You can sleep! At the heart of these computations are networks of neurons.
threeSim is a three-neuron simulator. Your brain has around 86 billion neurons, so this model is a bit simplified. However, you will see that just three neurons with fixed connections have a wide range of potential activity patterns. With this simulator, you can design your own neurons or select from a prefabricated set. You can modify synaptic strengths and time constants. You can begin to have an intuitive feel for the fundamental mechanisms of brain function.
This app is aimed at advanced undergraduate, graduate, and medical students, as well as professional scientists who want to know more about neural circuit interactions. If you don't fit that mold - Welcome!
Experiment and Learn
- Drive a postsynaptic neuron with excitatory synaptic input
- Temporally summate synaptic potentials
- Drive a postsynaptic neuron with inhibition
- Make neurons fire in particular patterns based on variations in synaptic input and intrinsic excitability
Under the hood
The cellular models used in threeSim are based on Izhikevich (2003). The simple, yet powerful, integrate-and-fire model enables realistic simulation of neurons with much less computational complexity than other biophysical models of neurons, such as the classic Hodgkin-Huxley models.
In addition to the neuronal models described by Izhikevich (2003), there is also a "neuron" type called "CC" for current clamp. I found when working with the model that it was sometimes difficult to show relatively simple synaptic phenomena, such as temporal summation of postsynaptic potentials, to naive users. This was primarily due to the intrinsic electrical membrane properties that are so well simulated by the Izhikevich model. I therefore created the CC type as a stripped-down model that reset the u to 0 and v to -45 mV after each millisecond, therefore effectively clamping the voltage-dependent currents to 0 and letting the synaptic potentials dominate.
Currently, no questions have been asked. You could be the first! Ask a support question.