Effects on gene expression in HSPCs. (A) LD score regression shows enrichments of heritability in regions with accessible chromatin in HSPC subpopulations (red nuance indicates –log₁₀P value). (B) cis-eQTLs at PPM1H, ENO1, RERE, and ITGA9 identified RNA-seq data for CD34⁺ cells from 155 blood donors (supplemental Table 10). Data are residual fragments per kilobase of transcript per million mapped reads (FPKM) values after correction for 10 expression principal components. Wedges indicate directions of effects on blood CD34⁺ cell levels for the same variant. (C) Candidate gene expression in single-cell messenger RNA-seq data from 35 582 blood and bone marrow mononuclear cells¹⁷ (supplemental Figures 9 and 10). ARHGAP45 was not represented in this data set. (D-E) Average expression of candidate genes across different cell clusters in single-cell Cellular Indexing of Transcriptomes and Epitopes by Sequencing data for 4905 lineage-negative CD34⁺ cells from adult bone marrow.¹⁸ In panel D, the 4905 cells have been clustered according to RNA-seq pattern (supplemental Figure 11). In panel E, the cells were instead clustered by gating using the sequence counts for the tags derived from antibodies to the CD38, CD45RA, CD90, CD123, and CD10 cell surface markers, as indicated at the upper edge of the heatmap (supplemental Figure 12). B, B cells; Baso, basophil; CD4, CD4⁺ T cells; CD8, CD8⁺ T cells; CLP, common lymphoid progenitors; Cyc, cycling cells; DC, dendritic cells; ERY, erythroid progenitors; GMP, granulocyte-monocyte progenitors; LMPP, lymphoid-primed multipotent progenitors; Mono, monocyte; PC, plasma cells; PreB, pre–B cells; Neut, neutrophil; NK, natural killer cells.