An In Silico Framework for the generation, simulatation and analysis of multi-scale neuron-network models
1 Max Planck Institute for Neurobiology of Behavior
2 Howard Hughes medical Institute
Brain functions are implemented by a complex interplay between mechanisms that span across several spatial and temporal scales, such as connectivity and activity at network scales, morphology and physiology at cellular scales, down to the expression and distributions of ion channels and synapses at subcellular scales. It is hence an enormous challenge to dissect the interplay between subcellular, cellular and circuit mechanisms that can account for in vivo observed activity patterns. Here we introduce ‘In Silico Framework’ (ISF) for the systematic and tractable exploration of mechanisms within and across scale that could account for in vivo observed activity patterns. Key features of ISF are the generation of millions of mechanistically detailed models, all of which are (1) equally well consistent with the empirical observations at each respective scale, (2) capture empirically observed variability within and across scales, and (3) thereby reveal distributions of multi-scale models that can in principle account for the in vivo observations, including variations from trial-to-trial and cell-to-cell thereof. Extending beyond previous in silico approaches that reported models of single scale mechanisms, ISF provides a publicly available set of tools for the generation, simulation and analysis of multi-scale models that predict neuron-network interactions underlying in vivo observations.