Researchers from Carnegie Mellon University, the University of Pittsburgh, and the University of Washington have mapped how 3D genome architecture changes in brain cells affected by Alzheimer's disease, linking genome folding to gene activity and tissue organization using single-cell technology, spatial mapping, and a deep learning model called Hicformer. By analyzing postmortem prefrontal cortex tissue with GAGE-seq and integrating it with spatial transcriptomics, the team connects chromosome structure to disease-related gene programs across cell types. Hicformer combines DNA sequence, genome-folding features, and local 3D contacts to predict cell-type–specific gene activity, enabling a multi-scale view of how structural changes relate to pathology. The findings indicate increased compartment mingling and weakened regulatory contacts, which may contribute to reduced synaptic function and microglial changes in Alzheimer’s brains. Related studies also show age-related shifts in genome regulation and immune-cell remodeling in aging, reinforcing the idea that genome architecture is a fundamental regulatory layer in neurodegeneration. Overall, researchers frame Alzheimer's as involving higher-order chromatin reorganization in addition to amyloid and tau, with implications for identifying new therapeutic targets.