Statistical analysis for identification of responder signature. (A) Heat map of top 100 probes that significantly differentiate responders (KG1, EOL-1, NOMO-1, HNT-34, FKH-1, and GDM-1) from nonresponders. Red, high expression; yellow, low expression. (B) Workflow of the statistical validation using German AMLCG 1999 clinical trial data as test data. The top 100 responder probes identified from the cell line study were used to retrieve relevant information from the test data. Using patient overall survival (OS) time, Cox proportional hazards survival analysis and Kaplan-Meier analysis were performed. (C) Fourteen genes in the test data were found to be significantly associated with patient OS time. (D) High expression of CD25 or GUCY1A3 (i) and low expression of EIF5A2 or MARCH3 (ii) were found to be significantly associated with shorter overall survival time. MOS, median of overall survival time. (E) Kaplan-Meier plot shows the high expression score from the combined signature significantly predicted poor patient survival time in the German AMLCG 1999 clinical trial data (P = 2.71e−7). MOS of “NA” indicates that more than 50% of the patients in the “low score” group survived until the end of the study.
