Skip to main content
. 2020 Jul 29;10:12761. doi: 10.1038/s41598-020-69691-8

Figure 5.

Figure 5

Validation based on independent DNMT3A-mutant AML patients. (A) Gene mutation based validation. Kaplan-Meier curves for an independent cohort of 208 DNMT3A-mutated AML with bone marrow transplantation from the German-Austrian AML Study Group. For each of these patients, the most similar DNMT3A-mutated AML patient of the TCGA cohort was determined by counting mismatches between the corresponding gene mutation profiles. Each patient was assigned to the short-lived or to the long-lived group based on the class label of the most similar TCGA patient (short-lived: red, 79 patients; long-lived: blue, 129 patients). Log-rank test for short- vs. long-lived: P < 0.003. (B) Gene expression based validation. Kaplan-Meier curves for an independent cohort of 63 DNMT3A-mutated AML patients from the University Hospital of Ulm that were also part of the German-Austrian AML Study Group. For each of these patients, correlations between its signature gene expression profile with the average short-lived and long-lived signature gene expression profiles of the DNMT3A-mutated AML patients from TCGA were computed. Each patient was assigned to the short-lived or to the long-lived group based on the maximum of both correlations (short-lived: red, 43 patients; long-lived: blue, 20 patients). Log-rank test for short- vs. long-lived: P < 0.03.