Fig. 4
From: Neuronal methylome reveals CREB-associated neuro-axonal impairment in multiple sclerosis
Meta-analysis of cohorts and comparison of DNA methylation with transcriptional changes. a Volcano plot illustrating differences in M-values between multiple sclerosis (MS) cases and non-neurological controls (adjusted P value < 0.05) from meta-analysis of cohorts 1 and 2. b Correlation between effect sizes (ΔM) of cohorts 1 and 2 for DMPs identified in meta-analysis (adjusted P value < 0.05) revealed predominant same direction of changes in the two cohorts. Regression line, P value (p) and Spearman’s rank correlation coefficient (r) are given. c Top canonical pathways of DMPs (adjusted P value < 0.05) identified in meta-analysis. Significance is represented as − log10(B-H, p value) with the red line showing threshold of significance after adjustment using Benjamini-Hochberg (B-H) correction. d Schematic representation of CREB-signalling pathway associated with DMPs from meta-analysis, with differentially methylated genes between cases and controls depicted in blue. e Association of 5mC-DMPs (left) and 5hmC-DhMPs (right) with gene expression data (RNA-seq) in bulk NAWM vs control WM [19]. Proportions (percentage) of number of upregulated and downregulated genes are depicted in red and blue, respectively (the dotted line represents the expected proportion). P values (p) generated with the Chi-square test are given for DMPs located in gene body or promoter-associated features (TSS1500, TSS200, 5′UTR and 1stExon) (full data in Additional file 8). f Scatter plots illustrating association between 5mC-DMPs (purple, left) and 5hmC-DhMPs (orange, right) methylation changes (Δβ-value) from cohort 2 (adjusted P value < 0.05) with gene expression data (RNA-seq) in bulk NAWM vs control WM [19]. The number of samples used in cohorts 1 and 2 equals to n = 12 and n = 17, respectively