This talk will highlight recent efforts to generate and analyze spatially resolved molecular datasets to better understand structure–function relationships in the human brain, particularly in the context of complex brain disorders. While single-cell and single-nucleus sequencing approaches have rapidly advanced our ability to define molecularly distinct cell populations, these methods often lack the spatial and circuit context necessary to interpret how cells interact within intact brain tissue. I will describe integrative strategies that combine spatial transcriptomics, single-cell genomics, and data-driven computational approaches to define molecularly distinct spatial domains within human brain regions, map cell–cell and circuit-level interactions across these domains, and identify enrichment of disease-associated molecular profiles in specific cellular and spatial contexts. Across examples from cortical and subcortical circuits, these approaches provide a framework for understanding how molecular heterogeneity is organized in space and how this organization may confer selective vulnerability in neuropsychiatric and neurodegenerative disease.