Neural circuit dynamics across a range of spatiotemporal scales endow the brain with specialized computational capabilities that subserve human cognition and behavior. Technological advances and big data initiatives in recent years have revolutionized our understanding of the brain’s multi-scale architecture. Yet there remains a major disconnect in linking large-scale dynamics of networked neural systems to their underlying circuit mechanisms. Biophysically-based computational modeling of neural systems provides a uniquely powerful framework for mechanistically linking these levels of analysis. This dissertation develops an extensible computational framework that integrates multi-modal brain data with neural systems modeling. We leverage this approach to link regional physiological differences in brain microcircuitry to the large-scale specialization of brain function. We further demonstrate how this modeling framework can generate and test predictions for the large-scale functional impacts of molecular perturbations, with relevance to psychiatry.
Thesis Advisor: John Murray (john.murray@yale.edu)