Extended Data Fig. 6: Single cell RNA sequencing analysis of fresh diagnostic samples.
a-b, Analysis of the biological states, processes and lineages underpinning the intercellular transcriptional heterogeneity amongst diagnostic cells from either Pt1 or Pt2 using the bioinformatics tool scran70 to decompose gene expression heterogeneity into technical noise and a biologically relevant component. A list of genes ranked by their biologically meaningful variance was then used as input for enrichment analysis with the fGSEA R-package using the Hallmark pathway gene sets (MSigDB) and lineage markers defined in71. The importance of each given signature in explaining variation amongst untreated cells is reflected in the normalized enrichment score (NES) reported next to the signature itself. P-values are obtained using 2-sided permutation test52 with 10,000 permutations. All shown processes have a Benjamini-Hochberg FDR-ajusted p-value <0.05. c, Heatmap showing the expression of gene sets associated with different lineages and differentiation stages of the hematopoietic hierarchy. Each line of the map represents an individual diagnostic cell from either Pt1 or Pt2. d, Bar plots visualizing the percentage of individual diagnostic cells from either Pt1 or Pt2 assigned to their closest normal hematopoietic cell lineage genes (signature with the highest Z-score average). The analysis is based on the expression of known markers32,53.
