Fig. 1: Extensive and consistent gene expression and chromatin accessibility across 25 human brain regions.

a Schematic representation of the study design. Postmortem samples were dissected from 25 brain regions across ForeBr, MidDien, BasGan, and HinBr. Nuclei were subjected to fluorescence-activated nuclear sorting (FANS) to yield neuronal (NeuN+) and non-neuronal (NeuN-) nuclei, followed by ATAC-seq and RNA-seq profiling. We also performed whole-genome sequencing for each individual. Sd represents standard deviation. Schematic was created using BioRender (https://biorender.com). Fullard, J. (2021) BioRender.com/h19m968, Fullard, J. (2021) BioRender.com/p82x857, and Fullard, J. (2021) BioRender.com/w09a902. b The 25 brain regions and abbreviations (anatomy dissection in Supplementary Fig.Ā 1). c Chromatin accessibility profiles merged from the neuronal broad brain regions around brain region-specific marker genes. d Pairwise statistical dissimilarity (quantified based on the proportion of true non-null tests, Ļ1) across different brain regions of neuronal and non-neuronal cells in the two assays. e the shared correlation structures (canonical correlation vectors, CC) between ATAC-seq and RNA-seq are separated by cell type (top, CC1-2) and brain regions (bottom, CC3-4). f Cell type enrichment determined by one-tailed Fisherās exact test in brain region-specific gene (for DEG)18 and OCR (for DAC)16 sets. Groups with a low number of significant differential results were masked and colored gray. Neu neuron. pyramidal SS somatosensory pyramidal cells. OPC oligodendrocyte progenitor cells. Odds ratio (OR).āĀ·ā: Nominally significant (Pā<ā0.05); ā+ā: significant after FDR (Benjamini & Hochberg) correction (FDRā<ā0.05). Source data is provided as a Source Data file.