(a) PESCA library plasmid map. ITR, inverted terminal repeats; GRE, gene regulatory element; pr, HBB minimal promoter; int, intron; GFP, green fluorescent protein; WPRE, Woodchuck Hepatitis Virus post-transcriptional regulatory element; BAR, 10-mer sequence barcode associated with each GRE; pA, polyadenylation signal. (b) Library complexity plotted as distribution of the abundance of the 861 barcodes and 287 GREs in the AAV library. Barcodes and GREs were binned by number of sequencing reads attributed to each barcode or GRE within the library. (c) Transcript count per nucleus (n = 32,335 nuclei). Sequencing libraries were prepared with or without PCR-enrichment for viral transcripts. PCR enrichment resulted in a 382-fold increase in the number of recovered viral transcripts (p=0, Mann-Whitney U-test, two-sided) to an average of 15.6 unique viral transcripts per nucleus. Displayed as Log10(Count+1). (d) t-SNE plot of 32,335 nuclei from V1 cortex of two animals. Colors denote main cell types: Exc (Excitatory neurons), Pv (PV Interneurons), Sst (SST Interneurons), Vip (VIP interneurons), Npy (NPY Interneurons), Astro (Astrocytes), Vasc (Vascular-associated cells), Micro (Microglia), Olig (Oligodendrocytes), OPCs (Oligodendrocyte precursor cells). (e) Marker gene expression across cell types. Color denotes mean expression across all nuclei normalized to the highest mean across cell types. Size represents the fraction of nuclei in which the marker gene was detected. (f) Dot plot with each dot representing one GRE (n = 287). The values on each axis represent the Log2 SST fold-enrichment calculated for each GRE based on two of the three barcodes paired with that GRE - barcode one on the x-axis, and barcode three on the y-axis. Blue line indicates linear fit with 95% confidence intervals (shaded) (r = 0.55, p<2.2 × 10−16, Pearson’s correlation). Color gradient indicates the average enrichment between the two barcodes. (g) Pairwise Pearson correlation between the enrichment values calculated from three sets of barcodes associated with 287 GREs for experimental data (Exp. Data, r = 0.52 ± 0.05, p<2.2 × 10−16, Pearson’s correlation) and after random shuffling of enrichment values (Shuffled Data, r = 0 ± 0.06). (h) GREs ranked by average barcode expression specificity for SST interneurons across three barcodes. Shading indicates the minimal and maximal specificity calculated by analyzing each of the three barcodes associated with a GRE. Blue indicates the five top hits that also passed a statistical test for SST interneuron enrichment (FDR-corrected q < 0.01). (i) Expression of the top five hits: GRE12, GRE19, GRE22, GRE44, GRE80. For each GRE, expression values are split into two animals, and, for each animal, into the three barcodes associated with that GRE. Color denotes mean expression across all nuclei normalized to the highest mean across cell types. Size represents the fraction of nuclei in which the marker gene was detected. (j) Mean expression of GRE12, GRE19, GRE22, GRE44, and GRE80 across cell types. Error bars, s.e.m.