Fig. 1: Overview of the single-nuclei dual-omic atlas of the human developing retina.

A The study design of this work. Samples were collected from either (1) whole retina at PCW 8, or (2) macula and periphery of the same retina from PCW 10 to PCW 23. Subsequently, with a total of 28 samples from 14 human embryos and fetuses, gene expression and open chromatin profiling from the same nuclei was performed using the 10X Chromium sn-dual-omic ATAC + Gene Expression technology (Supplementary Data 1). The bottom panel shows analysis workflow diagrams. B UMAP of RNA-Seq data colored by annotated major classes. C UMAP of ATAC-Seq data colored by annotated major classes. D Heatmap of the top 50 differentially expressed genes across each annotated major class (Supplementary Data 3). Each row is a nucleus and each column is a marker gene. E Heatmap of ATAC gene scores for the differentially expressed genes (same genes and nuclei with the same order as Fig. 1D) grouped by annotated major class. F UMAP of RNA-Seq data colored by annotated subclasses. G UMAP is separated by ___location and clock time. Dots were colored by annotated major class, and gray dots are background nuclei that are not from the corresponding time and ___location pair. H Line graphs of cell maturation scores against sample age (PCW) for each major class, colored by the macula and periphery. The cell maturation score represents the degree of a cell’s similarity to cells in the corresponding adult data. I Violin plots of previously published photoreceptor-maturation gene expression²⁸ between macula and periphery. To compare gene expression values temporally, two-sided overestimates variance t-test⁵⁹ was applied to different PCW groups among 47,446 photoreceptors. The Benjamini-Hochberg procedure was applied. J Line graphs of average cell age for each major class, colored by macula and periphery. Cells were down-sampling equally by PCW groups, and cell age was defined as sample’s age. To calculate means and error bars, 41,984 RGCs, 9532 HCs, 9324 cones, 26,384 ACs, 38,167 rods, 20,996 BCs, and 6553 MGs were used. Error bars represent the 95% confidence intervals from two-tailed t-tests. The Benjamini-Hochberg procedure was applied. Source data are provided as a Source Data file. Panel A was created with BioRender.com and released under a Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International license (https://creativecommons.org/licenses/by-nc-nd/4.0/deed.en).