Group meeting brief: June 3, 2014

Justin presented his simulation efforts on networks using integrate-and-fire neurons with membrane potential $V$ describe by the ODE $$C\frac{V}{dt} = -g_\mathrm{L} \left(V - E_\mathrm{L}\right) + I_\mathrm{syn}\left(t\right)+I_b + s\left(t\right)$$ where $C$ is membrane capacitance, $g_\mathrm{L}$ is leaking conductance, $E_\mathrm{L}$ is leaking reverse potential, $I_b$ is constant background current, and $s\left(t\right)$ is a Gaussian white noise. The synaptic transmission follows TUM dynamics which results in active neural transmitter fraction $Y$ and the synaptic current $$I_\mathrm{syn}= A \sum_{j} Y_j \left(V - V_\mathrm{rev}\right)$$ where $V_\mathrm{rev}$ is reverse potential for synaptic ion channels and $A$ is the constant synaptic weight for all synapses. Justin considered different networks with various mean degrees as well as degree distributions. He showed that the spiking activity, specifically, the power spectrum of spiking rates (measured with 5ms time bins) remains the same as long as the average synaptic input $\left\langle I_\mathrm{syn} \right\rangle$ is the same.

PS: While there are some questions to be answered regarding the influence of network structures on the oscillatory activity level of the spikes, Justin is approaching the problems using a newly acquired technique which combines power of python and c++.