(A) Heat map of bulk RNA-seq data showing the most highly variable 1000 genes (see Methods, maximum transcripts per million ≥10) within wildtype (WT1, WT2, and WT3) and CEP83−/− (KO1, KO2, and KO3) clones at day 7 of differentiation. Unbiased hierarchical clustering of clones separates CEP83−/− and WT transcriptomes. (B) Principal component analysis (PCA) of WT (WT1, WT2, WT3) and CEP83−/− (KO1, KO2, KO3) cells at day 7 using the average gene expression of the top highly variable 1000 genes in bulk RNA sequencing data. The % variation explained by each PCA axis is indicated in brackets. PCA eigenvalues indicate that the principal components, Dim 1 (52%) and Dim 2 (20.8%), account for 85.3% of the expression differences. Dim 1 separates the KO1 sample from the other samples, while Dim 2 separates experiment 1 (WT1, WT2, WT3) and (KO1, KO2, KO3).