Fig. 1: Multiscale brain organization, and workflow for model simulation and evaluation.

a First column displays a schematic diagram of different brain scales from microscopic to macroscopic level. Second column shows functions at different brain scales. Global dynamics quantified by metastability and E-I homeostasis are observed at macroscopic and mesoscopic scales, whereas, glutamate-GABA kinetics can be observed at microscopic level. Third column presents the working principles of the computational construct. The MDMF model placed at each node of an anatomical brain network (top tile) are optimally tuned by the two parameters (glutamate, GABA) across lifespan aging while maintaining E-I balance. b Overview of the entire pipeline for model fitting and evaluation. Empirical structural connectivity (SC) is constrained to the MDMF model that generates simulated FCs. Two conditions, i.e., maximizing metastability and minimizing FC distance between empirical and simulated FCs, are used to estimate optimal Glutamate and GABA values of for individual participants. Based on the optimal parameter values, simulated FCs are generated for individual participants, while the critical firing rates are maintained. The graph theoretic metrics of segregation and integration from the simulated FCs are measured and compared against the empirical FC properties, which provides an independent qualitative validation for the robustness of our results.