The quest for multiscale brain modeling

Computational models of neurotransmission at cerebellar synapses unveil the impact on network computation

“…In conclusion, the properties of neurotransmission and intrinsic electroresponsiveness determined experimentally in the cerebellum have been translated into accurate biophysical models, which, in turn, provide a new view of how the cellular and subcellular organization of neurons and synapses contribute to function. Indeed, while a bewildering number of molecular properties await functional explanation and system integration, simulations of detailed computational models can suggest new directions for experimental investigations and provide hints for understanding neuropathologies ( D’Angelo and Jirsa, 2022 ). For example, alterations in synaptic activity in brain diseases affecting the cerebellum, such as ataxia, dystonia, autism, and paroxysmal kinesigenic dyskinesia ( LeDoux et al, 1998 ; Rinaldo and Hansel, 2010 ; Brown and Loew, 2012 ; Won et al, 2013 ; Power and Empson, 2014 ; Zhang et al, 2015 ; Zhang and Sudhof, 2016 ; White and Sillitoe, 2017 ; Soda et al, 2019 ; Binda et al, 2021 ), remain largely to be elucidated; therefore, their translation into models would require investigation of structural changes, receptors, and biochemical pathways.…”

“…These results have revealed a previously undisclosed complexity and variety of mechanisms, requiring a further understanding of their functional implications through advanced computational models. Cellular and synaptic biophysics can be translated into mathematical equations ( Koch, 1984 ; Segev, 1998 ; Brodland, 2015 ) to generate models of neurons and networks, thus adding a new dimension to the computational investigation of neurotransmission ( D’Angelo and Jirsa, 2022 ). A relevant case is the cerebellar circuit, the synapses of which exhibit a variety of properties.…”

“…The extensive application of 3D morphological reconstructions ( Guo et al, 2022 ), immunochemistry, patch-clamp recordings, calcium imaging, genomics, and proteomics has allowed the creation of atlases, such as the Allen Brain Atlas ( Gilbert, 2018 ), which summarizes the knowledge about brain neurons and circuits along with the distribution of synaptic receptors and molecules, both in rodents and humans. Computational models can integrate this large set of experimental observations to reconstruct a variety of synaptic receptor kinetics (AMPA, NMDA, and GABA) and neurotransmission properties ( D’Angelo and Jirsa, 2022 ) ( Box 1 ).…”

Editorial: Neuroinformatics of large-scale brain modelling

“…The second main theme in this Research Topic is, as noted, improving connections between micro, meso, and macro levels of analysis, which is represented by the articles from Siu et al, Layer et al, and Maith et al. In the first of these, Maith et al introduce a blood oxygenation-dependent (BOLD) "monitor" (i.e., empirical measurement process simulator) into the ANNarchy (Vitay et al, 2015) spiking neuron modeling library. This work can be seen as part of the major recent and growing trend in computational neuroscience toward "true" multiscale neural models that simultaneously capture experimentally observed patterns across several qualitatively different data types (D'Angelo and Jirsa, 2022). The contribution by Siu et al also has a focus on modeling of hemodynamic activity patterns, in this case resting-state and stimulus-evoked fMRI activity patterns in anesthetized mice.…”

Implementation of the Hindmarsh–Rose Model Using Stochastic Computing

“…Simulation of the brain is one of the big issues for this century. The brain being the most complex structure in the known universe, problems inherent to this goal are many and different [1][2][3][4]. The current proposals cover many different approaches from quantum computing [5] to neural networks [6], including simulation at the level of tissues [7], interactions with different parts [8], or descriptions of the communications at a higher level [9].…”