The clinical impact of the balance between PD and ID signatures. (A-B) A heat map (left) depicting the H3K27ac PD (A) and ID (B) signatures. Signal intensities are presented as row z scores. The S_PD and S_ID signature enrichment scores for each sample are indicated below. A box plot (right) compares the S_PD or S_ID values between U-CLL and M-CLL. Statistical significance was assessed using 2-tailed Student t tests. (C) HRs for the multivariate Cox regression model including S_PD and S_ID as predictors of TTFT. Both scores were standardized as z scores for easier comparison and interpretation. (D) Definition of the balance score S_B. To illustrate the concept, this is represented as a quotient, despite the actual formula being S_PD – S_ID, effectively representing the PD-to-ID ratio, because these scores are determined from log-transformed data.²⁹ (E) Permutation test evaluating the χ² test statistic of the balance score when rearranging the labels in the clinical data. The histogram of the null distribution and a red vertical line indicating the observed statistic with the source data are plotted. (F) Comparison of Cox regression models using TTFT as the end point variable. The χ² test statistic and P value from the likelihood ratio test (LRT) are presented for each comparison. ∗∗∗∗P < .0001; ∗∗P < .01. CI, confidence intervals.