Correlation of proteome and transcriptome data sets. (A) Log₂ fold-change at protein level (x-axis) vs RNA level (y-axis). Significant changes at RNA (padj <.05 in green), protein (padj <.01 in blue), and both levels (in red) are highlighted. Correlation coefficient R² is shown for significant hits at both levels. (B) KEGG pathways overrepresented in hits significant at protein level and not significant at RNA level (blue) or vice versa (green). Fold enrichment and P value at protein level are shown. (C) Correlation of GSEA of KEGG and Hallmark gene sets on protein and RNA level in LSC vs blast comparisons. Correlated: unidirectional enrichment on both levels; anticorrelated: divergent enrichment on protein and RNA level, respectively. Relative distribution of metabolic and nonmetabolic gene sets per correlated or anticorrelated gene sets. (D) NES of KEGG and Hallmark gene sets at protein (x-axis) vs RNA level (y-axis). Metabolic gene sets are highlighted in red. (E) Selection of metabolic gene sets enriched in LSCs at protein level and not at RNA level. (F) Correlation of GSEA of KEGG and Hallmark gene sets on protein and RNA level in LSC vs HSPC comparisons. For legend, see panel C.