TET-family of gene dysfunction in myelodysplastic syndromes. (A) Bar plot showing 5-hmC levels (percent of 5-hmc/5-mC) in our internal cohort of patients (n = 81). Levels of statistical significance were assessed by Mann-Whitney U test. *P < .05, ***P < .0001. Pink quadrant indicates median level of expression in the WT cohort. On the right, pie charts representing percentage of patients with TET2 dysfunction (72%) and, among those, cases with TET2 mutations (60%). (B) Scatter plots of TET2 messenger RNA (mRNA) expression levels (log2_CPM). Left panel shows fractionated CD34⁺ cells of HC and MDS obtained from the Gene Expression Omnibus (accession number GSE63569) (P = .07). Middle panel shows a population of healthy controls (HC, n = 64) and our entire study cohort (MDS/MPN, n = 780). Right panel shows a breakdown of our study cohort (pure deletion4q24 = 13, TET2 mutant [MT] = 203, TET2, WT = 535, IDH MT = 29). Levels of statistical significance were assessed by Mann-Whitney U test. *P < .05, ***P < .0001. Pink quadrant indicates 25% lower percentile level of expression in the HC cohort. (C) Pie charts showing the percentages of patients in our cohort defined as low expressors and the breakdown of causes leading to TET2 dysfunction. (D) Scatter plot showing TET3 mRNA expression levels (log2_CPM) in a cohort of healthy controls (HC, n = 64), all cohort (MDS/MPN, n = 780), and TET2 WT (n = 577). Levels of statistical significance were assessed by Mann-Whitney U test. *P < .05, ***P < .0001. (E) TET3 vs TET2 mRNA expression (log2_CPM) in TET2 WT (n = 577). (F) Linear regression analyses of TET2 and TET3 mRNA expression (log2_CPM) vs TET2 variant allele frequency (n = 203). Variant allele frequency was expressed as percentage.