Analysis, tuning and implementation of neuronal models simulating Hippocampus dynamics

This work focuses on methods, algorithms and software for computational neuroscience, a research field where several computing strategies are intensively adopted for simulating real biological scenarios. This topic has a strong intersection with the Computer Science to build up frameworks that reproduce neuronal behaviours, coming from laboratory experiments and experiences. A key role is played by the simulation of synaptic mechanisms and neuronal dynamics that regulate the activity of the neurons. The model formalization requires long and deep steps in order to properly tune, through numerical simulations, a large number of biological parameters. These tasks, needed to validate a computational model, represent a well-known critical issue in terms of computing resources, both in time and in memory allocation. The main target of this PhD Thesis is the study of Hippocampus brain region, devoted to the acquisition of new memory (i.e., storage) and retrieval of previously acquired (i.e., recall). The dissertation starts with an introduction to the biological context and with the state of the art on some single cell and network models. Then, the formalization and the implementation of computational schemes that reproduce typical neuronal phenomena are deeply investigated. More in detail, the depolarization block of a CA1 pyramidal neuron and the effects of the CREB protein activity increasing in a CA1 microcircuit are investigated. For these topics, sequential and parallel code packages, published on the well-known neuroscience repository ModelDB, are implemented. Finally, as a case of study the acquired know-how on the modelling of cell and network dynamics was adopted for reproducing user behaviours in a cultural heritage scenario. Here, the main idea is to measure the interest of an artwork spectator by means of models able to capture the context and the visitor behaviours