The cell-type specific connectivity of the local cortical network explains prominent features of neuronal activity

In the past decade, the cell-type specific connectivity and activity of local cortical networks have been characterized experimentally to some detail. In parallel, modeling has been established as a tool to relate network structure to activity dynamics. While the available connectivity maps have been used in various computational studies, prominent features of the simulated activity such as the spontaneous firing rates do not match the experimental findings. Here, we show that the inconsistency arises from the incompleteness of the connectivity maps. Our comparison of the most comprehensive maps (Thomson et al., 2002; Binzegger et al., 2004) reveals their main discrepancies: the lateral sampling range and the specific selection of target cells. Taking them into account, we compile an integrated connectivity map and analyze the unified map by simulations of a full scale model of the local layered cortical network. The simulated spontaneous activity is asynchronous irregular and the cell-type specific spontaneous firing rates are in agreement with in vivo recordings in awake animals, including the low rate of layer 2/3 excitatory cells. Similarly, the activation patterns evoked by transient thalamic inputs reproduce recent in vivo measurements. The correspondence of simulation results and experiments rests on the consideration of specific target type selection and thereby on the integration of a large body of the available connectivity data. The cell-type specific hierarchical input structure and the combination of feed-forward and feedback connections reveal how the interplay of excitation and inhibition shapes the spontaneous and evoked activity of the local cortical network.