Extended Data Fig. 1. A multimodal molecular cell-type atlas of the MOp.

a, Anatomical location of the mouse MOp in the Allen Mouse Brain Common Coordinate Framework (CCFv3) in 3D and in representative sagittal and coronal sections. b–d, Documentation of MOp samples collected at the Allen Institute (b), the Broad Institute (c) and the Salk Institute (d). Each panel shows a diagram of coronal brain slices and dissected regions for transcriptomic (scRNA-seq and snRNA-seq) and epigenomic (snATAC and snmC-seq2) data samples based on the Allen Mouse Brain Common Coordinate Framework (CCF). Nissl-stained images in d show the posterior face of tissue slices (600 μm thickness). e, Number of cells and median number of unique sequenced DNA or RNA fragments per cell in each of the nine single-cell transcriptomic and epigenomic datasets. The squares show the extrapolated total library size based on the sequence duplication rate. f, Number of cells in each of the major cell classes (glutamatergic excitatory, GABAergic inhibitory neurons and non-neurons) of each dataset. Differences in cell-type sampling strategy, including the use of cell sorting to enrich neurons, affect the relative number of neurons and non-neuronal cells. Datasets include cells from the following numbers of mice (Supplementary Table 1): scRNA SMART: n = 28 male, 17 female; scRNA 10x v3 A: n = 3 male, 3 female; scRNA 10x v2 A: n = 3 male; snRNA SMART: n = 8 male, 2 female; snRNA 10x v3 B: n = 5 male, 6 female; snRNA 10x v2: n = 2 male, 1 female; snRNA 10x v3 A: n = 1 female; snmC-seq2 and snATAC-seq: n = 2 replicates, each pooled from 6 to 30 male mice. g, NeMO Analytics (nemoanalytics.org) visualization and analysis environment for the BICCN mouse molecular mini-atlas. Screenshot of NeMO Analytics showing multi-omic results for glutamate decarboxylase 2 (Gad2), a marker gene in inhibitory neurons. The web portal has the following features: (1) search box for gene names; (2) indicator of the gene viewed; (3) expandable species-specific functional annotation; (4) link-outs to additional resources for the selected gene; (5–7) interactive visualizations of each BICCN dataset, displayed in a ‘standalone’ box showing gene expression and cell clustering on integrated UMAP coordinates. Additional data exploration options for each of the datasets are available via the drop-down menu at the upper right corner of the NeMO Analytics dataset titles. (8) An embedded Epiviz interactive workspace to visualize scATAC-seq and sncMethyl-seq datasets in a linear browser view (8a), here showing the average ATAC and % CG methylation at the Gad2 locus (8c, 8d) as well as in each major cluster of glutamatergic and GABAergic neurons (8b, 8e, 8f). Epigenomic data are also available at http://epiviz.nemoanalytics.org/biccn_mop, and instructions for setting up and extending the Epiviz workspaces are available at http://github.com/epiviz/miniatlas. h, Brainome epigenomics portal (https://brainome.ucsd.edu/BICCN_MOp). The portal shows single-base resolution epigenomic and transcriptomic data (snmC-seq2, snATAC-seq, scRNA-seq and snRNA-seq) using the AnnoJ browser. Drop-down menus allow the user to select groups of cells (for example, excitatory, inhibitory and MGE-derived, among others), modalities (mCG, mCA, ATAC, scRNA, snRNA and enhancers) and display options. A Cell Browser allows visualization of scatter plots and heat maps of groups of genes across data modalities.