Skip to main content
. 2019 Sep 12;146(19):dev174722. doi: 10.1242/dev.174722

Fig. 5.

Fig. 5.

Increased transcript variation and differentiation upon loss of DNMT1. (A) Violin plot of single cell methylation data, where each dot represents the average CpG methylation level per cell. Cells were collected for scRNA-seq after 0, 2 and 8 days of doxycycline treatment. (B) Dimensionality reduction of day 0, 2 and 8 single cells (dots) using t-distributed stochastic neighbor embedding (t-SNE) and hierarchical clustering (bottom right) of the averaged expression profiles for in silico-sorted ES cell populations. (C) Fraction of cells classified into four categories (ES cell, dEN, dME, dEC) following 0, 2 and 8 days of doxycycline treatment. (D) Inter-sample (top) and intra-sample (bottom) density distribution of all pairwise cell-cell distances for in silico-sorted ES cells at day 0, 2 and 8. (E) Box plots of gene expression dispersion distribution at all genes, ES cell markers, and WT low dispersion genes for sorted ES cell populations at day 0, 2 and 8. (F) Differentially expressed genes (right; rows) for sorted population of ES cells at day 0, 2 and 8 (columns). Genes are separated into six gene sets [left: 100 (n=337), 101 (n=36), 110 (n=1301), 011 (n=349), 010 (n=139) and 001 (n=644)], where 1 or 0 indicates high or low expression for the respective condition (day 0, 2 and 8). (G) Functional enrichment analysis for the gene sets defined in F against the REACTOME database.