Abstract
Purpose :
Single-cell sequencing has dramatically transformed molecular profiling of tissues and organs. We recently created an integrated, multimodal reference atlas of the mammalian central nervous system's most accessible part, the retina. This atlas comprises a transcriptomics profile of approximately 2.4 million cells and an open chromatin profile of about 400K nuclei from 55 donors, revealing over 110 distinct cell types. Our goal is to explore this large dataset to gain insights into retinal biology.
Methods :
scVI is used to integrate the scRNA-seq datasets while scGlue integrates snRNA-seq and snATAC-seq cells. Deep learning algorithms such as Pando and SCIENIC+ were used to identify gene regulatory networks, key transcription factors, and their binding motifs. We examined associations between gene expression and chromatin accessibility to link regulatory elements to target genes. A community-based annotation effort by an expert panel improved cell annotation accuracy. Additionally, raw and processed data are available in public databases like CELLxGENE, UCSC browser, and HCA data portal for easy access and exploration.
Results :
In addition to the transcriptomic profiles of over 110 cell types in the retina, more than 600K open chromatin regions have been identified. Systematic characterization of cis-regulatory elements for individual cell types has linked the enhancers to their targeted genes. Integrative analysis of transcriptomics and epigenomics data has identified key gene regulatory networks and transcription factors underlying the functions of each cell type. Moreover, changes in cell proportion, gene expression, and chromatin openness have been observed across different genders and ages. Finally, by combining genomics and GWAS data, cell types strongly associated with diseases, as well as putative causal variants and genes associated with various diseases, have been identified.
Conclusions :
This study, accessible via interactive web browsers, stands as the most detailed and comprehensive cell atlas of the human retina available so far. In-depth analysis of this extensive dataset has yielded numerous insights into retinal biology and associated diseases. As a component of the Human Cell Atlas project, this resource establishes a solid foundation for advancing research into the understanding of retinal biology and related disorders.
This abstract was presented at the 2024 ARVO Annual Meeting, held in Seattle, WA, May 5-9, 2024.