Biological constraints for parameter values of large-scale biologically plausible human Neuroscience models
Description
This dataset contains 1.) time constants of excitatory and inhibitory neocortical neurons, collated from the NeuroElectro database, 2.) spike thresholds of excitatory and inhibitory neurons, collated from the NeuroElectro database, and 3.) posterior distributions of parameters of large-scale biologically plausible models of human Neuroscience data, estimated using likelihood-free inference (LFI) methods to fit these models to Magnetoencephalography (MEG) resting-state data (N=75) in Williams et al. (2023), NeuroImage paper 4.) Data used to perform Prior Predictive Checks in Williams et al. (2023), NeuroImage paper, 5.) Data used to perform Posterior Predictive Checks in Williams et al. (2023), NeuroImage paper. We used pooled versions of the time constants and spike threshold values in this dataset along with other biological constraints, to set prior distributions of parameters of large-scale biologically plausible models of human Neuroscience data. We then used LFI methods to estimate posterior distributions of these parameters from MEG resting-state data (N=75) - we have also shared the posterior distributions in this dataset. Examples of parameters for which we have shared posterior distributions include strengths of connections within and between excitatory and inhibitory neuronal populations, time constants of excitatory and inhibitory neuronal populations, and firing thresholds of excitatory and inhibitory neuronal populations. These posterior distributions could be used by other research groups to set prior distributions of parameters of their biologically plausible models, in both experimental and modeling studies.
Files
Institutions
Categories
Funding
Academy of Finland
Helsinki Institute for Information Technology
Department of Science & Technology, India
Sigrid Juselius Foundation
Finnish Center for Artificial Intelligence