Fig. 1

Experimental design, quality assessment, and unsupervised clustering. A Human islet processing and data generation scheme. B Data analysis workflow. C Number of genes per cell/nucleus, number of reads per cell/nucleus, and mitochondrial gene ratio comparison between scRNA-seq /snRNA-seq data. D Number of genes and number of reads per cell/nucleus between snRNA-seq data with and without intronic reads. Statistical significance was tested using a Wilcoxon rank-sum test. E Average gene expression correlation among different human islet cell/nuclei preparations (n = 3 adult human islet donors). F Average gene expression correlation between RNA sequencing type (scRNA/snRNA). G Venn diagram of genes detected in both scRNA-seq and snRNA-seq analysis of the three human islet samples, UMI > 20. H Unsupervised clustering of scRNA-seq and snRNA-seq integrated data with Louvain resolution of 0.8. I. Dimensional reduction plot grouped by RNA sequencing type