Cortical networks are characterized by inhomogeneities in structure and dynamical properties [1] and excitatory cortical cells were shown to form fine-scale subnetworks [2, 3]. These properties seem to play an important role in their functioning [4]. In a previous study [5], we presented a simulation study of a balanced state network where the interplay of different forms of synaptic plasticity lead to the emergence of long-tailed weight-distributions from an initially homogeneous state. Additionally, we observed that a small fraction of excitatory neurons that we call driver neurons develop predominantly strong outgoing synapses. Apart from having a strong outgoing connections, they are also distinguished by elevated firing rates and form subnetworks with an increased degree of connectivity, similar to some cortical subnetworks found in recent experiments [2,3]. Here, we investigate analytically and numerically the nature of the interaction of the different plasticity mechanisms that allow for the self-organized development of driver neurons.

[1] S. Song, P. J. Sjostroem, M. Reigl, S. Nelson, and D. B. Chklovskii. Highly nonrandom features of synaptic connectivity in local cortical circuits. PLOS Biology, 3(3):e68 (2005).
[2] S. Lefort, C. Tomm, J-C. Floyd Sarria, and C. C. H. Petersen. The excitatory neuronal network of the C2 barrel column in mouse primary somatosensory cortex. Neuron, 61(2):301-16 (2009).
[3] L. Yassin, B. L. Benedetti, J-S. Jouhanneau, J. A. Wen, J. F. A. Poulet, and A. L. Barth. An embedded subnetwork of highly active neurons in the neocortex. Neuron, 68(6):1043-50 (2010).
[4] Gyoergy Buzsaki and Kenji Mizuseki. The log-dynamic brain: how skewed distributions affect network operations. Nat Rev Neurosci, 15(4):264-278 (2014).
[5] Felix Effenberger, Anna Levina, and Juergen Jost. On the influence of inhibitory stdp on balanced state random networks. BMC Neuroscience, 14(Suppl 1):P200 (2013).