Extended Data Fig. 7: The Primitive versus Mature axis governs ex vivo drug sensitivity.

A) Volcano plot depicting associations between Primitive versus GMP axis (PC1) and ex vivo drug sensitivity from the BEAT-AML (Tyner et al 2018) drug screen (n = 202 patients), and between PC1 and PC2 and ex vivo drug sensitivity from Lee et al 2018 (n = 30 patients). B) Unsupervised clustering of 30 primary AML patients from Lee et al on the basis of ex vivo sensitivity to 159 drugs. Drug sensitivity values (scaled negative AUC) are depicted for the top drugs corresponding to each patient cluster. Red denotes sensitivity while blue denotes lower sensitivity. C) GSEA analysis with gene signatures derived from the Primitive vs Mature axis in normal and malignant hematopoiesis, performed on genes ranked by differential expression between the two drug response clusters from (B). D) Correlations of AML gene expression scores with ex vivo drug sensitivities from two drug screens (Tyner et al, n = 202; Lee et al, n = 30) performed on primary AML samples. E) Ex vivo drug sensitivity to Venetoclax (n = 114 from Tyner et al) and Azacytidine (n = 30 from Lee et al) of primary patient samples grouped into ‘High’ or ‘Low’ based on median splits of patient scores for each AML gene expression score. Significance was evaluated through two-sided Wilcoxon rank-sum tests. Box plots indicate the range of the central 50% of the data, with the central line marking the median. Whiskers extend from each box to 1.5*(interquartile range).