A central focus in contemporary neuroscience research is the mapping and modelling of connectivity and activity dynamics in large-scale brain networks. As the resolution, coverage, and availability of neural data increases rapidly, neuroinformatics techniques are playing an increasingly important role in this scientific enterprise. Large-scale brain modelling is the methodologically-defined sub-field of computational neuroscience that is focused on simulations of either whole-brain activity at a coarse-grained (meso/macro) spatial scale, or activity in select neural subsystems at a fine-grained (micro) spatial scale and high level of detail. Neuroinformatics tools employed in large-scale brain modelling come in the form of organizing concepts and software infrastructure that facilitate the core work of computational modelling and data analysis.
Recent advances in this area have included the development and use of large-scale detailed databases on brain anatomy, connectivity, gene expression, and activity recordings in the service of large-scale brain simulations; techniques for databasing, scientific unit testing, and validation of computational models across large parameter and simulated data feature spaces; and efforts to capture empirical phenomena at multiple spatial scales of organization within unified multi-scale brain simulations. In many cases the neuroinformatics and architectural solutions developed as part of this work are in themselves of general methodological interest to researchers, but are communicated secondarily to the principal neuroscientific research questions. Moreover, despite significant progress in these and other domains, there has to date been no major attempts to survey the diverse contributions of neuroinformatics to large-scale brain modelling as a whole. One of the aims of this Research Topic is therefore to present the computational neuroscience community with a series up-to-date summaries of and perspectives on the state of the art, highlighting valuable new approaches, as well as current deficiencies and knowledge/infrastructure gaps.
This Research Topic will document the various ways in which neuroinformatics approaches are being applied in large-scale brain modelling, informing readers on both established practices and emerging techniques. We seek Original Research, Review, Mini-Review, Hypothesis and Theory, Perspective, and Opinion articles that cover, but are not limited to, the following topics:
• Ontologies, systems, and tools for definition and specification of neural models
• New approaches to parameter optimization, parameter space exploration, and systematic tracking of simulation behaviour across parameter combinations
• Informing neural models with genetic and multi-omic data from large-scale databases and individual patients/subjects
• Systematic computational modelling studies on large numbers of subjects, and/or using large-scale open-access datasets (HCP, ABCD, etc.)
• ‘Hybrid’ modelling schemes that combine mean-field with spiking network models
• ‘Hybrid’ approaches to defining connectivity in large-scale brain models (e.g. supplementing tractography with microscopy data for higher-resolution subcortical connectivity structure)
• Simulations using BigBrain
• ‘High-density’ (large number of regions; small parcels) connectome-based neural mass modelling
• Other neuroinformatics challenges and solutions in large-scale brain simulations
• Comparisons between detailed spiking/morphological simulations and neural mass model simulations
• Comparisons between models based on high-resolution and low-resolution Allen atlas connectivities
A central focus in contemporary neuroscience research is the mapping and modelling of connectivity and activity dynamics in large-scale brain networks. As the resolution, coverage, and availability of neural data increases rapidly, neuroinformatics techniques are playing an increasingly important role in this scientific enterprise. Large-scale brain modelling is the methodologically-defined sub-field of computational neuroscience that is focused on simulations of either whole-brain activity at a coarse-grained (meso/macro) spatial scale, or activity in select neural subsystems at a fine-grained (micro) spatial scale and high level of detail. Neuroinformatics tools employed in large-scale brain modelling come in the form of organizing concepts and software infrastructure that facilitate the core work of computational modelling and data analysis.
Recent advances in this area have included the development and use of large-scale detailed databases on brain anatomy, connectivity, gene expression, and activity recordings in the service of large-scale brain simulations; techniques for databasing, scientific unit testing, and validation of computational models across large parameter and simulated data feature spaces; and efforts to capture empirical phenomena at multiple spatial scales of organization within unified multi-scale brain simulations. In many cases the neuroinformatics and architectural solutions developed as part of this work are in themselves of general methodological interest to researchers, but are communicated secondarily to the principal neuroscientific research questions. Moreover, despite significant progress in these and other domains, there has to date been no major attempts to survey the diverse contributions of neuroinformatics to large-scale brain modelling as a whole. One of the aims of this Research Topic is therefore to present the computational neuroscience community with a series up-to-date summaries of and perspectives on the state of the art, highlighting valuable new approaches, as well as current deficiencies and knowledge/infrastructure gaps.
This Research Topic will document the various ways in which neuroinformatics approaches are being applied in large-scale brain modelling, informing readers on both established practices and emerging techniques. We seek Original Research, Review, Mini-Review, Hypothesis and Theory, Perspective, and Opinion articles that cover, but are not limited to, the following topics:
• Ontologies, systems, and tools for definition and specification of neural models
• New approaches to parameter optimization, parameter space exploration, and systematic tracking of simulation behaviour across parameter combinations
• Informing neural models with genetic and multi-omic data from large-scale databases and individual patients/subjects
• Systematic computational modelling studies on large numbers of subjects, and/or using large-scale open-access datasets (HCP, ABCD, etc.)
• ‘Hybrid’ modelling schemes that combine mean-field with spiking network models
• ‘Hybrid’ approaches to defining connectivity in large-scale brain models (e.g. supplementing tractography with microscopy data for higher-resolution subcortical connectivity structure)
• Simulations using BigBrain
• ‘High-density’ (large number of regions; small parcels) connectome-based neural mass modelling
• Other neuroinformatics challenges and solutions in large-scale brain simulations
• Comparisons between detailed spiking/morphological simulations and neural mass model simulations
• Comparisons between models based on high-resolution and low-resolution Allen atlas connectivities
