Supercomputing facilities are becoming increasingly available in neuroscience, predominantly for simulating the dynamics of electrical activity propagating through neuronal circuits. On today's most advanced supercomputers, large-scale models can be simulated with up to a billion of neurons. However, merely further increasing the number of neurons and synapses will not be sufficient to create biologically realistic full-scale brain models. Brain networks are sparsely connected but, importantly, dynamically rewire throughout life, thereby permanently maintaining a critical and functional balance of neuronal activity. Existing large-scale models of brain networks, however, have no detailed local and global anatomical connectivity and moreover, anatomical connectivity in these models is fixed, with plasticity merely arising from changes in synaptic strengths. Finding the “right” layout of local and global connections is thus a major challenge for designing large- and full-scale brain models. The focus of this Research Topic is therefore on approaches that bring high-performance computing in neuroscience together with the latest experimental research in neuroanatomy. The first aspect of this Research Topic will focus on large-scale models, high-performance simulation tools, and novel hardware and data-base techniques that try or allow to capture the highest possible degree of anatomical detail. The second aspect will include approaches that introduce forms of anatomical plasticity that make detailed full-scale connectivity adaptable, e.g. to changes in experience or during learning and memory consolidation. A third aspect will capture novel concepts in the self-organization and self-repair of large-scale and full-scale brain networks, with potential applications for neural development and the diseased brain, respectively. With the aim to present work that goes beyond the current stage of high-performance computing in neuroscience, we welcome computational and technical contributions with strong emphasis on or applicability to neuroanatomical research, as well as experimental contributions that discuss important ways to improve large-scale brain models. We will consider all types of submission, from original research to more review-type articles. Especially, we encourage experimental neuroscientists to submit perspectives or opinions on new trends and discoveries relevant to large-scale brain models.
Supercomputing facilities are becoming increasingly available in neuroscience, predominantly for simulating the dynamics of electrical activity propagating through neuronal circuits. On today's most advanced supercomputers, large-scale models can be simulated with up to a billion of neurons. However, merely further increasing the number of neurons and synapses will not be sufficient to create biologically realistic full-scale brain models. Brain networks are sparsely connected but, importantly, dynamically rewire throughout life, thereby permanently maintaining a critical and functional balance of neuronal activity. Existing large-scale models of brain networks, however, have no detailed local and global anatomical connectivity and moreover, anatomical connectivity in these models is fixed, with plasticity merely arising from changes in synaptic strengths. Finding the “right” layout of local and global connections is thus a major challenge for designing large- and full-scale brain models. The focus of this Research Topic is therefore on approaches that bring high-performance computing in neuroscience together with the latest experimental research in neuroanatomy. The first aspect of this Research Topic will focus on large-scale models, high-performance simulation tools, and novel hardware and data-base techniques that try or allow to capture the highest possible degree of anatomical detail. The second aspect will include approaches that introduce forms of anatomical plasticity that make detailed full-scale connectivity adaptable, e.g. to changes in experience or during learning and memory consolidation. A third aspect will capture novel concepts in the self-organization and self-repair of large-scale and full-scale brain networks, with potential applications for neural development and the diseased brain, respectively. With the aim to present work that goes beyond the current stage of high-performance computing in neuroscience, we welcome computational and technical contributions with strong emphasis on or applicability to neuroanatomical research, as well as experimental contributions that discuss important ways to improve large-scale brain models. We will consider all types of submission, from original research to more review-type articles. Especially, we encourage experimental neuroscientists to submit perspectives or opinions on new trends and discoveries relevant to large-scale brain models.
