Week 1: The Neuron Model

There are two distinct types of neurons within the Antennal Lobe: projection neurons (PNs) and local neurons (LNs). These neurons are relatively similar, though they vary slightly in their behaviors.


For starters, PNs are mainly excitatory whereas LNs are inhibitory. Excitatory neurons push other neurons closer to their firing threshold, whereas inhibitory neurons push neurons further away from reaching their firing threshold.The timescale of PN excitation is relatively short (~2 ms), whereas the timescale of LN inhibition is both short (~2 ms) and long (~800 ms).


Furthermore, PNs exhibit a unique property known as SK current. Essentially, whenever a PN neuron fires an action potential, it inhibits itself through its SK current. Little is known about the details of the SK current, making it an interesting topic for investigation later in the project.


To actually model the neurons, we required a mathematical model to approximate a neurons voltage at any given time. We decided to use a conductance based model, which, in essence, determines how permeable the neuron membrane is to various ions in order to determine where the voltage of the neuron lies.


Composing the conductance model are a variety of different signals that either increase or decrease the total voltage. Signals such as excitation and sensory input increase the voltage, whereas signals such as fast inhibition, slow inhibition, leak conductance, and, for PNs only, SK current all serve to decrease the total voltage.


Finally, we had to mathematically model the signals themselves. Input was modeled using a poisson random process, where a background noise rate indicated the frequency with which noisy signals would come in to the neurons. SK currents were modeled using a short rise time (~50 ms) resembling a sinusoidal curve, followed by an exponential decay on a longer time scale (~400 ms). All other inputs were modeled with instantaneous rise times for simplicity, and exponential decays with time scales mimicking their realistic time scales. With all of these considerations in place, by the end of the week we had fully functioning independent neurons. Next week we will look into grouping neurons together to form distinct glomeruli.

Speak Your Mind