Fig. 2: Global proteome analysis shows stability through progression and groups cases with poor outcome.

a Descriptive summary of the cohort for proteome analysis. The bar plot represents the total number of each cytogenetic subtype and the donut plots represent the age (top) and sex (bottom) of the patients. b Hierarchical clustering of 3907 variable proteins represented by the relative log2FC (protein intensity/median protein intensity). The color bars indicate sample type (bottom) and leukemic blast percentage (top). c The major B-ALL clusters were selected for in-depth characterization. Hierarchical clustering of 935 proteins based on log2FC defined seven sample clusters (horizontal) and five protein clusters (vertical). The other color bars indicate main cytogenetic subgroup (second from the bottom), followed by clinically assigned risk group (SR = standard risk, HR = high risk), and current survival status on top. The five most significant GO terms for each cluster of proteins were selected for visualizations. The annotation to the right of each cluster of proteins is the summary of the top significant terms. Bars represent the adjusted p-value for each GO term. d Kaplan Meier survival curve with up to 5 year follow-up data for B-ALL clusters (n = 38) grouped by clinically assigned risk group (SR = standard risk, HR = high risk), ns = not significant by logrank test for trend. Black tick marks on the survival curve represent data that has been censored due to follow-up data <5 years. e Kaplan Meier survival curve with up to 5 year follow-up data for B-ALL samples grouped by proteome cluster for clusters with >4 samples (n = 30). Significance assigned by logrank test for trend. Black tick marks on the survival curve represent data that has been censored due to follow-up data <5 years.