Frontiers in Computational Neuroscience (Aug 2016)
Multifactorial modeling of impairment of evoked gamma range oscillations in schizophrenia
Abstract
Despite a significant increase in efforts to identify biomarkers and endophenotypic measuresof psychiatric illnesses, only a very limited amount of computational models of these markersand measures has been implemented so far. Moreover, existing computational models dealingwith biomarkers typically only examine one possible mechanism in isolation, disregarding thepossibility that other combinations of model parameters might produce the same networkbehaviour (what has been termed ’multifactoriality’). In this study we describe a step towardsa computational instantiation of an endophenotypic finding for schizophrenia, namely theimpairment of evoked auditory gamma and beta oscillations in schizophrenia. We explorethe multifactorial nature of this impairment using an established model of primary auditorycortex, by performing an extensive search of the parameter space. We find that singlenetwork parameters contain only little information about whether the network will show impairedgamma entrainment and that different regions in the parameter space yield similar networklevel oscillation abnormalities. These regions in the parameter space, however, show strongdifferences in the underlying network dynamics. To sum up, we present a first step towards anin silico instantiation of an important biomarker of schizophrenia, which has great potential forthe identification and study of disease mechanisms and for understanding of existing treatmentsand development of novel ones.
Keywords
