Skip to main content
. 2022 Sep 22;16:884245. doi: 10.3389/fninf.2022.884245

FIGURE 1.

FIGURE 1

Reproduction and validation of BBP S1 cell types in NetPyNE. (A) 3D reconstructions of 4 pairs of m-type example neurons visualized using the NetPyNE graphical user interface: inhibitory cells L1_DLAC and L4_DBC (red), and excitatory cells L23_PC and L6_TPC_L4 (blue). (B–E) Somatic membrane potential of the neurons in (A) under current clamp with amplitude 120% of the neuron firing threshold. NetPyNE results (red) compared to the original BBP model results (blue). For L4_DBC_bIR cells (D) we used a deterministic version of the BBP stochastic potassium channel (StochKv) resulting in divergent results; using same deterministic channel in BBP (BBPdet, blue dotted line) restores the match to NetPyNE results. (F,G) Comparison of BBP and NetPyNE firing rate and time to first spike in response to current-clamp with amplitude 0.1 nA during 2 s for each cell type. Only some cell types with the StochKv show differences. (H) Backpropagating action potential in a L5_TTPC cell. (I) Post synaptic potentials (PSPs) of one dendritic connection with 5 excitatory (blue circles), 10 excitatory (red squares), and 20 inhibitory (cyan diamonds) synapses. (J) Comparison BBP and NetPyNE mean firing rate for all m-type populations. Due to the small number of cells with StochKv (3.6%), NetPyNE population firing rates closely match those of BBP.