a, Proportion of all PN, PVALB, and SST interneuron subtypes in MOs, SSp, and VISp cortices based on MERFISH data. b, Violin plots showing the laminar distribution of selected PVALB and SST subtypes across three cortical regions. c, Bar plots showing the layer-by-layer composition of PVALB and SST subtypes in each region. d, Boxplots showing the ratio of interneurons (INs) to PNs and the contribution of PVALB and SST subtypes to the total interneuron pool across regions (Center line, median; box limits, 25th and 75th percentiles; whiskers, 1.5xIQR). Same dataset as in Fig. 1d, n = 9 (MOs), 13 (SSp), 9 (VISp) ROIs. Two-sided Wilcoxon rank-sum test, n.s. not significant, P ≥ 0.05; * P < 0.05; ** P < 0.01; *** P < 0.001. Exact P values are provided in Supplementary Table 3. e–h, Validation of regional interneuron densities by immunolabeling and genetic targeting. e, PVALB interneurons were labeled by immunostaining (n = 3 mice; P26-53). Note: This antibody-based method was chosen as it Is largely specific to PVALB interneurons, whereas Pvalb in situ hybridization (see Extended Data Fig. 7a) or Pvalb driver based genetic labeling, which also label L5 PT neurons in SSp. f, SST interneurons were visualized using genetic labeling. Comparable distributions were confirmed by in situ hybridization for Sst mRNA (n = 5 mice total across both methods; P26-61). g, Labeling of L4-targeting SST–Hpse interneurons with some off-target SST-Calb2 interneurons (n = 3 mice; P30-36); h, Labeling of L5b SST–Chrna2 interneurons (n = 5 mice; P34-P40). For e-h, Scale bars, 500 µm (main) and 100 µm (higher magnification of outlined regions on the right). Exact sample genotypes are provided in Supplementary Table 2. i, UMAP of P28 snRNA-seq data based on only the genes included in the MERFISH probe set, demonstrating that this gene set is sufficient to resolve the defined subtypes.