Publicação
Structural and field heterogeneity in synchronization dynamics on complex networks: application to neuronal networks
| Resumo: | The structural and dynamical properties of neuronal networks and other complex systems are emergent. Interaction between billions of neurons gives rise to collective phenomena, and the rules that shape neuronal networks lie hidden in structural correlations. This thesis aims to extend existing knowledge of synchronous activity in real directed networks, neuronal networks in particular, and thus advance the study of complex systems and address open questions in neuroscience. First, source nodes are shown to disrupt collective phenomena through feedforward connectivity into the strongly connected core of directed networks. This finding may further our understanding of neurodegenerative diseases, since the random destruction of neurons and synapses creates source nodes. These nodes lack incoming links, thus acting as an external field on downstream nodes in all networks with a bow-tie architecture. Second, synchronization and entrainment in a periodic external field are studied using the reduced Kuramoto model (KM). Analysis of the model reveals two distinct disrupted states, where oscillators are neither synchronized nor entrained, but instead fall in and out of phase with each other, while the group phase either oscillates or rotates relative to the external field. Oscillating and rotating states are shown to be topologically distinct, with characteristic winding numbers, and found to coexist in systems where different fractions of oscillators are exposed to local field phases and strengths. The impact of field heterogeneity and the distinct nature of disrupted states provide fundamental insight into the behavior of real systems such as the brain's suprachiasmatic nucleus (SCN). This thesis studies a core-shell model of the SCN based on the reduced KM and experimental parameters, and shows that the properties of disrupted states are in good agreement with experimental observations of circadian rhythm dissociation in the SCN under varying day length. In addition, the core-shell organization is shown to enable anticipation of future events over circadian cycles, and the model is modified to account for Aschoff's first rule. Third, and finally, a survey of the central nervous system of different animals shows that reciprocally connected pairs of nodes (RPCs) are a common motif with clear functional roles. However, the rules governing the formation and modification of synapses in RCPs remain an open problem. In the context of learning and memory, there is evidence that Hebbian synaptic plasticity can lead to the formation of new synapses in RCPs. Thus motivated, this thesis performs a statistical analysis of the synaptic multiplicity of connections in the chemical connectome of the C. elegans roundworm, an animal model of neuroplasticity. Structural correlations between synaptic multiplicity and the number of presynaptic neighbors is found to be compatible with Hebbian structural plasticity. The analysis also identifies sex-specific differences of unknown origin in the distribution of RCPs. |
|---|---|
| Autores principais: | Wright, Edgar António Policarpo |
| Assunto: | Complex systems Directed networks Bow-tie architecture Source nodes Kuramoto model Synchronization Entrainment Topological transition Winding number Field heterogeneity Suprachiasmataic nucleus Circadian rhythms Coreshell organization Dissociation Anticipation Aschoff's first rule Reciprocal synapses Synaptic multiplicity C. elegans Connectome Structural synaptic plasticity |
| Ano: | 2022 |
| País: | Portugal |
| Tipo de documento: | tese de doutoramento |
| Tipo de acesso: | acesso aberto |
| Instituição associada: | Universidade de Aveiro |
| Idioma: | inglês |
| Origem: | RIA - Repositório Institucional da Universidade de Aveiro |
| Resumo: | The structural and dynamical properties of neuronal networks and other complex systems are emergent. Interaction between billions of neurons gives rise to collective phenomena, and the rules that shape neuronal networks lie hidden in structural correlations. This thesis aims to extend existing knowledge of synchronous activity in real directed networks, neuronal networks in particular, and thus advance the study of complex systems and address open questions in neuroscience. First, source nodes are shown to disrupt collective phenomena through feedforward connectivity into the strongly connected core of directed networks. This finding may further our understanding of neurodegenerative diseases, since the random destruction of neurons and synapses creates source nodes. These nodes lack incoming links, thus acting as an external field on downstream nodes in all networks with a bow-tie architecture. Second, synchronization and entrainment in a periodic external field are studied using the reduced Kuramoto model (KM). Analysis of the model reveals two distinct disrupted states, where oscillators are neither synchronized nor entrained, but instead fall in and out of phase with each other, while the group phase either oscillates or rotates relative to the external field. Oscillating and rotating states are shown to be topologically distinct, with characteristic winding numbers, and found to coexist in systems where different fractions of oscillators are exposed to local field phases and strengths. The impact of field heterogeneity and the distinct nature of disrupted states provide fundamental insight into the behavior of real systems such as the brain's suprachiasmatic nucleus (SCN). This thesis studies a core-shell model of the SCN based on the reduced KM and experimental parameters, and shows that the properties of disrupted states are in good agreement with experimental observations of circadian rhythm dissociation in the SCN under varying day length. In addition, the core-shell organization is shown to enable anticipation of future events over circadian cycles, and the model is modified to account for Aschoff's first rule. Third, and finally, a survey of the central nervous system of different animals shows that reciprocally connected pairs of nodes (RPCs) are a common motif with clear functional roles. However, the rules governing the formation and modification of synapses in RCPs remain an open problem. In the context of learning and memory, there is evidence that Hebbian synaptic plasticity can lead to the formation of new synapses in RCPs. Thus motivated, this thesis performs a statistical analysis of the synaptic multiplicity of connections in the chemical connectome of the C. elegans roundworm, an animal model of neuroplasticity. Structural correlations between synaptic multiplicity and the number of presynaptic neighbors is found to be compatible with Hebbian structural plasticity. The analysis also identifies sex-specific differences of unknown origin in the distribution of RCPs. |
|---|