Fig. 3
From: scMET: Bayesian modeling of DNA methylation heterogeneity at single-cell resolution
Summary of changes in methylation patterns (mean and variability) for inhibitory and excitatory neurons. a Identifying differentially methylated (DM) features for the distal H3K27ac genomic context. Green and pink points correspond to features showing higher mean methylation in inhibitory (Inh+) and excitatory (Exc+) neurons, respectively. Labels highlight neuron marker genes that were used in [9] to define the different cell populations. Blue vertical dashed lines correspond to log-odds ratio threshold ψM=± log(2). Yellow horizontal dashed line is located at posterior evidence probability cut-offs defined by EFDR = 5%. b Representative heatmap of methylation rates (Yij/nij) across cells (columns) and features (rows) for the distal H3K27ac genomic context. Cells are grouped in excitatory and inhibitory classes. A set of randomly selected features is displayed. These are grouped according to the DM analysis output as: Exc+, Inh+, and no mean methylation difference. The color code represents features with low (0, green) and high (1, yellow) mean methylation level. Features with no CpG coverage are denoted with white color. c Identifying differentially variable (DV) features for the distal H3K27ac genomic context. Green and pink points correspond to features showing higher methylation variability in inhibitory and excitatory neurons, respectively. Blue dashed lines correspond to log-odds ratio threshold ψE=± log(1.5). Yellow dashed line is located at posterior evidence probability cut-offs defined by EFDR = 5%. For each feature, posterior estimates for the change in residual overdispersion parameter εj between excitatory and inhibitory neurons is plotted against the posterior tail probability of calling a feature as DV (left). For each feature, posterior estimates for mean methylation parameter μj is plotted against posterior estimates for the change in residual overdispersion parameter εj between excitatory and inhibitory neurons (right). d Example features that are called as being more variable in excitatory neurons. Left subplots show broad differences in methylation patterns. Right subplots show methylation patterns separately within each broad neuronal class. Each data point represents the methylation rate for a cell