• dentate gyrus
• computational model
• active dendrites
• adaprive exponential integrate and fire model
• pattern separation
• sparse activity

Hippocampus is known to be involved in numerous memory processes, including episodic and spatial memory. The hippocampal Dentate Gyrus (DG) is believed to be responsible for distinguishing between similar mnemonic events, a computation known as pattern separation. Granule cells, the principal neurons of the DG, are thought to perform pattern separation by minimizing the overlap between similar patterns. Granule cells are sparsely connected, fire scarcely due to strong attenuation of propagating synaptic currents, and integrate strong inhibitory signals, generating a "winner-takes-all" mechanism. Moreover, recent findings suggest that the dendrites of GCs are able to support local regenerative events, the so-called dendritic spikes, whose potential functional role remains unclear. In this thesis, we create a biologically constrained, computational model of the Dentate Gyrus network that consists of GCs and various interneurons (Basket cells, HIPP cells, CCK cells) simulated as integrate-and-fire models.
We used the model to assess the relative importance of the various characteristics of GC, focusing primarily on how active dendrites may influence pattern separation. Preliminary results predict that active dendrites facilitate pattern separation via decreasing the activity of Granule cells and thus increasing network sparsity. Further investigation is necessary to explain this unexpected finding, via a detailed biophysical model and/or in vitro dendro-somatic recordings.