All barcodes labelled in the global object as cardiomyocytes, fibroblasts and neural cells were selected for further subpopulation analyses. Additional cell population-specific filtering criteria were applied to nuclei as follows: cardiomyocyte counts (n_counts <12,500), genes (n_genes <4,000), mitochondrial genes (percent_mito <1%), ribosomal genes (percent_ribo <1%) and scrublet score (scrublet_score <0.25); FB mitochondrial genes (percent_mito <1%), ribosomal genes (percent_ribo <1%); neuronal cell genes (n_genes <4000), mitochondrial genes (percent_mito <1%), ribosomal genes (percent_ribo <1%). Total and CD45+ cells were excluded in the atrial and ventricular cardiomyocytes datasets and did not contribute to subpopulation analysis. No further filtering of FBs or neuronal cell total and CD45+ cells was applied. Cardiomyocytes and FBs were then further split into two groupings based on the region of origin: (1) left and right atrium, and (2) left and right ventricles, apex and interventricular septum.
Donor effects were aligned as described in step (1) above. For FB and neuronal cells, sources were aligned as described in step (3) above. Leiden clustering and UMAP visualization were performed for identifying subpopulations and visualization86 (link). Differentially expressed genes were calculated using the Wilcoxon rank sum test. Genes were ranked by score.
Donor effects were aligned as described in step (1) above. For FB and neuronal cells, sources were aligned as described in step (3) above. Leiden clustering and UMAP visualization were performed for identifying subpopulations and visualization86 (link). Differentially expressed genes were calculated using the Wilcoxon rank sum test. Genes were ranked by score.
Full text: Click here