Extended Data Fig. 8: Characterization of CLL-PD by proteomic, ex vivo drug response and bioenergetic profiling.
a, Correlations between CLL-PD to the protein levels of three MYC direct targets that are involved in cell proliferation: MCM4, NME1 and PAICS. Per-test P values and coefficients are from two-sided Pearson’s correlation tests (n = 46 samples). b, P values of associations between drug responses and F1 (IGHV), F2 (trisomy 12) and F4 (CLL-PD). P values are from ANOVA tests including F1, F2 and F4 as covariates. Dashed horizontal line indicates the threshold associated with a false discovery rate (FDR) of 5% (method of Benjamini and Hochberg). c, Scatter plots showing the correlations between cell viabilities after drug treatment (averaged over five concentrations tested) and the CLL-PD values. P values were from the same ANOVA test as shown in panel b. Only the drugs that showed significant correlations (5% FDR) are shown here. (panel b and c n = 190 samples): d, Scatter plots showing the associations of CLL-PD to the three bioenergetic features related to oxidative phosphorylation. Per-test P values and coefficients were from two-sided Pearson’s correlation tests (n = 136 samples). e, A heatmap plot showing the z-score of the expression values of proteins that are significantly correlated with CLL-PD (5% FDR, method of Benjamini and Hochberg). Samples (columns) are ordered by their CLL-PD values. The names of mitochondrial proteins are colored in red. f, The correlation between the CLL-PD values of 10 samples and their mitochondrial biomass, analyzed by MitoTracker staining. MitoTracker Green (ThermoFisher Scientific, M7514) was used according to the compound’s manual. P value and coefficient are from two-sided Pearson’s correlation tests. g, Correlations between CLL-PD and the expressions of two mitochondrial marker proteins, VDAC1 and HSPD1 (HSP60). Per-test P values and coefficients are from two-sided Pearson’s correlation tests (n = 46 samples).
