Skip to main content

Thank you for visiting You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

Human haematopoietic stem cell lineage commitment is a continuous process


Blood formation is believed to occur through stepwise progression of haematopoietic stem cells (HSCs) following a tree-like hierarchy of oligo-, bi- and unipotent progenitors. However, this model is based on the analysis of predefined flow-sorted cell populations. Here we integrated flow cytometric, transcriptomic and functional data at single-cell resolution to quantitatively map early differentiation of human HSCs towards lineage commitment. During homeostasis, individual HSCs gradually acquire lineage biases along multiple directions without passing through discrete hierarchically organized progenitor populations. Instead, unilineage-restricted cells emerge directly from a ‘continuum of low-primed undifferentiated haematopoietic stem and progenitor cells’ (CLOUD-HSPCs). Distinct gene expression modules operate in a combinatorial manner to control stemness, early lineage priming and the subsequent progression into all major branches of haematopoiesis. These data reveal a continuous landscape of human steady-state haematopoiesis downstream of HSCs and provide a basis for the understanding of haematopoietic malignancies.

Access options

Rent or Buy article

Get time limited or full article access on ReadCube.


All prices are NET prices.

Figure 1: Experimental strategy.
Figure 2: A stem and progenitor cell continuum precedes the establishment of discrete lineages at the CD34+CD38+ stage.
Figure 3: The LinCD34+CD38+ compartment consists of distinct lineage-restricted progenitors.
Figure 4: Characterization of LinCD34+CD38+ lineage-restricted progenitors.
Figure 5: Visualization of the HSPC continuum.
Figure 6: The direction of transcriptomic priming is quantitatively linked to functional lineage potential.
Figure 7: The degree of transcriptomic priming is quantitatively linked to multipotency and proliferative capacity.
Figure 8: Lineage commitment is a layered multi-step process.

Accession codes

Primary accessions

Gene Expression Omnibus


  1. 1

    Chao, M. P., Seita, J. & Weissman, I. L. Establishment of a normal hematopoietic and leukemia stem cell hierarchy. Cold Spring Harb. Symp. Quant. Biol. 73, 439–449 (2008).

    CAS  PubMed  Google Scholar 

  2. 2

    Morrison, S., Uchida, N. & Weissman, I. The biology of hematopoietic stem cells. Annu. Rev. Cell Dev. Biol. 11, 35–71 (1995).

    CAS  PubMed  Google Scholar 

  3. 3

    Kondo, M., Weissman, I. L. & Akashi, K. Identification of clonogenic common lymphoid progenitors in mouse bone marrow. Cell 91, 661–672 (1997).

    CAS  PubMed  Google Scholar 

  4. 4

    Akashi, K., Traver, D., Miyamoto, T. & Weissman, I. L. A clonogenic common myeloid progenitor that gives rise to all myeloid lineages. Nature 404, 193–197 (2000).

    CAS  PubMed  PubMed Central  Google Scholar 

  5. 5

    Doulatov, S. et al. Revised map of the human progenitor hierarchy shows the origin of macrophages and dendritic cells in early lymphoid development. Nat. Immunol. 11, 585–593 (2010).

    CAS  PubMed  PubMed Central  Google Scholar 

  6. 6

    Notta, F. et al. Isolation of single human hematopoietic stem cells capable of long-term multilineage engraftment. Science 333, 218–221 (2011).

    CAS  PubMed  Google Scholar 

  7. 7

    Notta, F. et al. Distinct routes of lineage development reshape the human blood hierarchy across ontogeny. Science 351, aab2116 (2016).

    PubMed  Google Scholar 

  8. 8

    Sun, J. et al. Clonal dynamics of native haematopoiesis. Nature 514, 322–327 (2014).

    CAS  PubMed  PubMed Central  Google Scholar 

  9. 9

    Busch, K. et al. Fundamental properties of unperturbed haematopoiesis from stem cells in vivo. Nature 518, 542–546 (2015).

    CAS  PubMed  Google Scholar 

  10. 10

    Perié, L., Duffy, K. R., Kok, L., de Boer, R. J. & Schumacher, T. N. The branching point in erythro-myeloid differentiation. Cell 163, 1655–1662 (2015).

    PubMed  Google Scholar 

  11. 11

    Haas, S. et al. Inflammation-induced emergency megakaryopoiesis driven by hematopoietic stem cell-like megakaryocyte progenitors. Cell Stem Cell 17, 422–434 (2015).

    CAS  PubMed  Google Scholar 

  12. 12

    Görgens, A. et al. Revision of the human hematopoietic tree: granulocyte subtypes derive from distinct hematopoietic lineages. Cell Rep. 3, 1539–1552 (2013).

    PubMed  Google Scholar 

  13. 13

    Adolfsson, J. et al. Identification of Flt3+ lympho-myeloid stem cells lacking erythro-megakaryocytic potential. Cell 121, 295–306 (2005).

    CAS  Google Scholar 

  14. 14

    Yamamoto, R. et al. Clonal analysis unveils self-renewing lineage-restricted progenitors generated directly from hematopoietic stem cells. Cell 154, 1112–1126 (2013).

    CAS  PubMed  Google Scholar 

  15. 15

    Naik, S. H. et al. Diverse and heritable lineage imprinting of early haematopoietic progenitors. Nature 496, 229–232 (2013).

    CAS  PubMed  Google Scholar 

  16. 16

    Paul, F. et al. Transcriptional heterogeneity and lineage commitment in myeloid progenitors. Cell 163, 1663–1677 (2015).

    CAS  Google Scholar 

  17. 17

    Wilson, N. K. et al. Combined single-cell functional and gene expression analysis resolves heterogeneity within stem cell populations. Cell Stem Cell 16, 712–724 (2015).

    CAS  PubMed  PubMed Central  Google Scholar 

  18. 18

    Trapnell, C. et al. The dynamics and regulators of cell fate decisions are revealed by pseudotemporal ordering of single cells. Nat. Biotechnol. 32, 381–386 (2014).

    CAS  PubMed  PubMed Central  Google Scholar 

  19. 19

    Shin, J. et al. Single-cell RNA-Seq with waterfall reveals molecular cascades underlying adult neurogenesis. Cell Stem Cell 17, 360–372 (2015).

    CAS  PubMed  Google Scholar 

  20. 20

    Olsson, A. Single-cell analysis of mixed-lineage states leading to a binary cell fate choice. Nature 537, 698–702 (2016).

    CAS  PubMed  PubMed Central  Google Scholar 

  21. 21

    Theilgaard-Monch, K. The transcriptional program of terminal granulocytic differentiation. Blood 105, 1785–1796 (2005).

    PubMed  Google Scholar 

  22. 22

    Borregaard, N. Neutrophils, from marrow to microbes. Immunity 33, 657–670 (2010).

    CAS  PubMed  Google Scholar 

  23. 23

    Clark, M. R., Mandal, M., Ochiai, K. & Singh, H. Orchestrating B cell lymphopoiesis through interplay of IL-7 receptor and pre-B cell receptor signalling. Nat. Rev. Immunol. 14, 69–80 (2013).

    PubMed  PubMed Central  Google Scholar 

  24. 24

    Haghverdi, L., Büttner, M., Wolf, F. A., Buettner, F. & Theis, F. J. Diffusion pseudotime robustly reconstructs lineage branching. Nat. Methods 13, 845–848 (2016).

    CAS  Google Scholar 

  25. 25

    Hoppe, P. et al. Early myeloid lineage choice is not initiated by random PU.1 to GATA1 protein ratios. Nature 535, 299–302 (2016).

    CAS  Google Scholar 

  26. 26

    Fischbach, N. A. et al. HOXB6 overexpression in murine bone marrow immortalizes a myelomonocytic precursor in vitro and causes hematopoietic stem cell expansion and acute myeloid leukemia in vivo. Blood 105, 1456–1466 (2005).

    CAS  PubMed  Google Scholar 

  27. 27

    Iacovino, M. et al. HoxA3 is an apical regulator of haemogenic endothelium. Nat. Cell Biol. 13, 72–78 (2011).

    CAS  PubMed  Google Scholar 

  28. 28

    Chuikov, S., Levi, B. P., Smith, M. L. & Morrison, S. J. Prdm16 promotes stem cell maintenance in multiple tissues, partly by regulating oxidative stress. Nat. Cell Biol. 12, 999–1006 (2010).

    CAS  PubMed  PubMed Central  Google Scholar 

  29. 29

    Ito, K. & Suda, T. Metabolic requirements for the maintenance of self-renewing stem cells. Nat. Rev. Mol. Cell Biol. 15, 243–256 (2014).

    CAS  PubMed  PubMed Central  Google Scholar 

  30. 30

    Shojaei, F. et al. Hierarchical and ontogenic positions serve to define the molecular basis of human hematopoietic stem cell behavior. Dev. Cell 8, 651–663 (2005).

    CAS  PubMed  Google Scholar 

  31. 31

    Kataoka, K. et al. Evi1 is essential for hematopoietic stem cell self-renewal, and its expression marks hematopoietic cells with long-term multilineage repopulating activity. J. Exp. Med. 208, 2403–2416 (2011).

    CAS  PubMed  PubMed Central  Google Scholar 

  32. 32

    Frelin, C. et al. GATA-3 regulates the self-renewal of long-term hematopoietic stem cells. Nat. Immunol. 14, 1037–1044 (2013).

    CAS  PubMed  PubMed Central  Google Scholar 

  33. 33

    Hattangadi, S. M., Wong, P., Zhang, L., Flygare, J. & Lodish, H. F. From stem cell to red cell: regulation of erythropoiesis at multiple levels by multiple proteins, RNAs, and chromatin modifications. Blood 118, 6258–6269 (2011).

    CAS  PubMed  PubMed Central  Google Scholar 

  34. 34

    Novershtern, N. et al. Densely interconnected transcriptional circuits control cell states in human hematopoiesis. Cell 144, 296–309 (2011).

    CAS  PubMed  PubMed Central  Google Scholar 

  35. 35

    Signer, R. A. J., Magee, J. A., Salic, A. & Morrison, S. J. Haematopoietic stem cells require a highly regulated protein synthesis rate. Nature 509, 49–54 (2014).

    CAS  PubMed  PubMed Central  Google Scholar 

  36. 36

    Friedman, A. D. Transcriptional control of granulocyte and monocyte development. Oncogene 26, 6816–6828 (2007).

    CAS  PubMed  PubMed Central  Google Scholar 

  37. 37

    Hystad, M. E. et al. Characterization of early stages of human B cell development by gene expression profiling. J. Immunol. 179, 3662–3671 (2007).

    CAS  PubMed  Google Scholar 

  38. 38

    Kurotaki, D. et al. IRF8 inhibits C/EBPα activity to restrain mononuclear phagocyte progenitors from differentiating into neutrophils. Nat. Commun. 5, 4978 (2014).

    CAS  PubMed  Google Scholar 

  39. 39

    Waddington, C. H. The Strategy of the Genes (Routledge, 1957).

    Google Scholar 

  40. 40

    Brock, A., Chang, H. & Huang, S. Non-genetic heterogeneity—a mutation-independent driving force for the somatic evolution of tumours. Nat. Rev. Genet. 10, 336–342 (2009).

    CAS  PubMed  Google Scholar 

  41. 41

    Huang, S. Non-genetic heterogeneity of cells in development: more than just noise. Development 136, 3853–3862 (2009).

    CAS  PubMed  PubMed Central  Google Scholar 

  42. 42

    Freud, A. G. & Caligiuri, M. A. Human natural killer cell development. Immunol. Rev. 214, 56–72 (2006).

    CAS  PubMed  Google Scholar 

  43. 43

    Essers, M. A. G. et al. IFNα activates dormant haematopoietic stem cells in vivo. Nature 458, 904–908 (2009).

    CAS  PubMed  Google Scholar 

  44. 44

    Picelli, S. et al. Smart-seq2 for sensitive full-length transcriptome profiling in single cells. Nat. Methods 10, 1096–1098 (2013).

    CAS  Google Scholar 

  45. 45

    Sasagawa, Y. et al. Quartz-Seq: a highly reproducible and sensitive single-cell RNA-Seq reveals non-genetic gene expression heterogeneity. Genome Biol. 14, R31 (2013).

    PubMed  PubMed Central  Google Scholar 

  46. 46

    Picelli, S. et al. Tn5 transposase and tagmentation procedures for massively scaled sequencing projects. Genome Res. 24, 2033–2040 (2014).

    CAS  PubMed  PubMed Central  Google Scholar 

  47. 47

    Wu, T. D. & Nacu, S. Fast and SNP-tolerant detection of complex variants and splicing in short reads. Bioinformatics 26, 873–881 (2010).

    CAS  PubMed  PubMed Central  Google Scholar 

  48. 48

    Anders, S., Pyl, P. T. & Huber, W. HTSeq–a Python framework to work with high-throughput sequencing data. Bioinformatics 31, 166–169 (2015).

    CAS  Google Scholar 

  49. 49

    Kharchenko, P. V., Silberstein, L. & Scadden, D. T. Bayesian approach to single-cell differential expression analysis. Nat. Methods 11, 740–742 (2014).

    CAS  PubMed  PubMed Central  Google Scholar 

  50. 50

    Velten, L. et al. Single-cell polyadenylation site mapping reveals 3’ isoform choice variability. Mol. Syst. Biol. 11, 812 (2015).

    PubMed  PubMed Central  Google Scholar 

  51. 51

    Jaitin, D. A. et al. Massively parallel single-cell RNA-seq for marker-free decomposition of tissues into cell types. Science 343, 776–779 (2014).

    CAS  PubMed  PubMed Central  Google Scholar 

  52. 52

    Van Dongen, S. Graph clustering via a discrete uncoupling process. SIAM J. Matrix Anal. Appl. 30, 121–141 (2008).

    Google Scholar 

  53. 53

    Csardi, G. & Nepusz, T. The igraph software package for complex network research. InterJournal Complex Systems, 1695 (2006).

    Google Scholar 

Download references


We thank C. Drumm for help with 3D graphics, K. Hexel, S. Schmitt, C. Felbinger and M. Eich from the DKFZ flow cytometry facility for flow cytometry support, the EMBL Genomics Core Facility for sequencing and R. Aiyar, A. Jones, M. Milsom and all members of HI-STEM and the Steinmetz group for helpful discussions on the manuscript as well as T. Schroeder and D. Löffler for initial discussions. This work was supported by the SFB873 funded by the Deutsche Forschungsgemeinschaft (DFG) (to C.L., M.A.G.E. and A.T.), the Dietmar Hopp Foundation (to M.A.G.E. and A.T.) and the US National Institutes of Health (P01 HG000205 to L.M.S.).

Author information




S.F.H., S.R., L.V., S.B. and C.H. performed the experiments. L.V. analysed the data, with conceptual input from S.F.H., S.R., L.M.S., M.A.G.E. and A.T., and analytical advice from W.H. S.I. and B.P.H. optimized genomics methods. C.L., E.C.B., D.N., T.B., W.-K.H. and A.D.H. obtained bone marrow aspirates. L.V., S.F.H., S.R., M.A.G.E., L.M.S. and A.T. jointly conceived and designed the study, and wrote the manuscript.

Corresponding authors

Correspondence to Andreas Trumpp or Marieke A. G. Essers or Lars M. Steinmetz.

Ethics declarations

Competing interests

The authors declare no competing financial interests.

Integrated supplementary information

Supplementary Figure 1 Flow cytometric display of the setup used for index-omics.

(a) Sorting was performed exclusively on the gates highlighted in red in the CD34 versus CD38 panel and all surface markers were indexed. Cells outside the green gate in the CD45RA versus CD90 panel were excluded retrospectively as they represented mature immune cells (not shown). Percentage of cells within each gate is indicated. Data from individual 1 is shown. (b) Distribution of cells subjected to single-cell RNA-Seq within classically defined gates1,2.

Supplementary Figure 2 Quality metrics of single-cell RNA-Seq.

(a) Bioanalyzer traces of amplified cDNA generated from single human HSPCs with the default smart-seq2 protocol (upper panel), QUARTZ-Seq (middle panel, applied to individual 2) and a modified version of smart-seq2 (lower panel, applied to individual 1, see methods). (b,c) Filtering of cells based on total read counts and number of genes expressed. The use of the modified smart-seq2 protocol (b) strongly decreased the dropout rate compared to the QUARTZ-Seq protocol (c). The large dropout rate in the QUARTZ-Seq protocol was due to the small volume used. (d,e) The number of genes per cell (d) and cell-cell correlation (e) for the two individuals compared to two other recent single-cell RNA-Seq data sets from the haematology field3,4. Box plots display median bar, first–third quantile box and 5th–95th percentile whiskers. n = 379 cells individual 1, n = 1,034 cells individual 2, n = 218 cells Individual 1, HSCs; n = 2,730 cells Paul et al. n = 1,058 cells Kowalczyk et al.. (f,g) The mean read count and variance of spike-ins (large black dots) and genes (small dots) were compared to identify genes whose biological noise exceeded technical variability (cyan dots)5. (h) The total RNA content of LinCD34+cells varies widely. (i) Cartoon describing the hypothetical effect of large variations in RNA amount in homogeneous populations. Two cells from the same population (red) display identical RNA concentrations for two genes, but differ in RNA amount by 10-fold. A third cell from a different population expresses the two sample genes at a different ratio but absolute high number. Following sequencing, the genes are more likely to be lost in the smaller cell, which cannot be reverted by normalization. (j) PCA performed on lymphoid (CLP) specific genes6 should clearly separate cells expressing the lymphoid surface marker CD10. However, without normalization cells are only arranged by read count (i). Standard normalization using a harmonic mean estimator of library size does not solve the problem (ii). Following normalization by Posterior Odds Ratios (POR, see Online Methods) a PCA performed on CLP specific genes clearly separates CD10+ and CD10 cells (iii).

Supplementary Figure 3 indeXplorer, a web-based GUI for exploring single-cell index-omics and index-culture data.

indeXplorer combines the capabilities of a FACS software with tools for the analysis of single cell transcriptomics data in a single graphical user interface. FACS and transcriptomics modules are tightly linked, allowing for example the display of gene expression or transcriptomic clusters on FACS scatter plots (a), differential expression testing of arbitrarily gated populations (b), as well as hierarchical clustering (c) and principal component analysis (d). indeXplorer further provides tools for gene list management, allows the user to download plots as publication-quality pdfs, and to store & restore sessions. On we provide a short interactive introduction into the use of indeXplorer.

Supplementary Figure 4 Unsupervised analyses of single-cell transcriptomics.

(a) Cluster stability analysis7 of the LinCD34+CD38 and LinCD34+CD38+ populations. For n = 500 repetitions, 66% of cells were randomly selected, clustering was performed and a consensus clustering was computed. The probability that clusterings obtained from random subsets of the data agree with the consensus is plotted on the y axis (box plot with median bar, first–third quantile box and 5th–95th percentile whiskers). (b) Gap-statistic (Gapk) of LinCD34+CD38 and LinCD34+CD38+ compartments. A maximum of Gapk indicates the statistically optimal cluster number8. (c) clustering obtained using ICGS9. 4 outlier cells in the LinCD34+CD38 compartment (left panel, blue bar) were characterized by a lower number of genes detected, but no coherent differences in gene expression (not shown). (d)–(f) Transcriptomic heterogeneity in the LinCD34+CD38 compartment. (d) > 10 principal components in LinCD34+CD38 exceed noise. (e) Principal components 2 and 5 of a PCA performed on combined data from both individuals. Loadings of all genes with annotated cell-cycle phase dependent gene expression patterns10 are shown in the right panel. Cell cycle associated genes are shifted compared to other genes on PC2 and arranged by peak time of gene expression on PC5. Scores of all LinCD34+CD38 cells are shown in the left panel. (f) Principal components 3 and 4. Loadings of all genes annotated as CD38+CD10+ ‘CLP’ or CD41+CD42+GP6+ ‘Mk’ specific6 are shown, demonstrating that PC3 and PC4 correlate with lymphoid versus megakaryocytic priming. Scores of all LinCD34+CD38 cells are shown in the left panel. (g) Principal components of LinCD34+CD38 cells are significantly correlated to surface marker expression. Data from individual 1 are shown. (h) Expression of neutrophil marker genes in relation to CD45RA and CD135. See also Main Fig. 4c. (i) Expression of cell cycle genes suggests that the CD10midFSC-Ahigh population is more actively cycling. (j) Ki67-Hoechst cell cycle analyses of IL7RCD9+ and IL7R+CD9 populations, corresponding to sB and lB respectively. (k) Cells from the transcriptomic Im cluster have intermediate CD38 expression and group with LinCD34+CD38 HSPCs in t-SNE analysis.

Supplementary Figure 5 Analyses using STEMNET.

(a) The similarity of every cell to each of the progenitor classes was computed by STEMNET (see methods), projected on a unit circle, and used to quantify the degree and direction of transcriptomic priming. Data from individual 2 is shown. (b) immunophenotypes highlighted on the STEMNET plot for individual 2. (c) CD38 surface marker expression highlighted on the STEMNET plot for individual 1. (d,e) Dual lineage primed cells, defined as cells with more than 25% priming in two directions, were highlighted on the STEMNET plot (d) or in a ternary plot depicting only priming in the Mk, Neutro, and Eo/Baso/Mast directions (e). (f) Rare IRF8+GFI1+ progenitors9 are not a typical intermediate stage between granulocytes and monocytes but appear displaced from developmental trajectories or are fully primed towards individual lineages. (g) Distribution of colony types observed in the index-culture experiment. Functionally bipotent cells are highlighted. (h,i) The transcriptomic lineage priming of immunophenotypic CMPs depends strongly on the gating strategy. Cells from the CMP gate (LinCD34+CD38+CD45RACD135+) were highlighted on the STEMNET plot (upper panels) or as ternary plots (lower panels). The effect of variations in the CD135 (h) and CD38 (i) gates are shown. P-values were calculated by kernel-density based tests comparing each population to CD49f+ HSCs. For CD49f + HSCs, n = 101 single cells; CMPs, default gate, n = 64; CMPs, relaxed CD38 gate, n = 164; CMPs, stringent CD38 gate, n = 24; CMPs, relaxed CD135 gate, n = 180.

Supplementary Figure 6 Simulation of data from alternative models of cell fate specification.

(a) To demonstrate the ability of STEMNET to identify subsequent binary branching events, we assumed a scenario where cells locate on developmental trajectories between universally defined branching points (left panel, see also methods). For each gene used by STEMNET as a marker specific to a given developmental endpoint, we reordered the expression data to parallel the developmental distance from that endpoint. The middle panel depicts exemplary the reordered expression of CSF3R, a neutrophil marker. Finally, we apply STEMNET to the reordered data set (right panel). (b) To simulate data using a more realistic noise level, we estimated the correlation between developmental distance and gene expression from the data for each gene (upper panels). We then reshuffled the expression values such that the correlation between marker gene expression and (simulated) developmental distance approximates the correlation estimated from the data (lower panels). (c) STEMNET on reshuffled data. (d) To simulate a scenario where HSCs pass through discrete progenitor cell types, cells were placed near branching points, data was simulated as described for panel (b), and STEMNET was applied to the reshuffled data. (e) Projection of single cell expression data into diffusion map space11.

Supplementary Figure 7 The quantitative link between index-omics and index-culture.

(a) Regression models used to estimate transcriptomic quantities from FACS surface marker expression. Model coefficients and the fraction of variance explained in a 10-fold cross validation scheme (R2) are shown. For genes marked with an asterisk, regression models were constructed on mRNA expression and applied to FACS surface marker expression. b, Linkage of the exact predicted direction of transcriptomic priming (for the cell types with robust colony forming abilities; Neutro, Ery, Mk) to the actual cell type composition of the ex vivo colonies. Illustration (left panel) and quantitative linkage (right panel) are shown. The exact direction of transcriptomic priming was estimated for each founder cell from index-culture based on regression models constructed on all surface markers and compared to the observed colony composition. (c) CD71 and KEL FACS marker and mRNA expression in relation to the degree of transcriptomic Ery/Mk priming and the percentage of Ery/Mk cells in the colony. (d,e) As an additional experimental measure of developmental plasticity, we cultured single HSPCs for 1 week, split the colony in four and determined the lineage outcome of the daughter colonies two weeks later. For several colonies, the lineage output varied significantly across daughters (e, P-values are from a chi-square test for independence). These colonies tended to derive from developmentally more primitive cells (d). P-value is from a Pearson product moment correlation test with n = 96 split-in-four experiments.

Supplementary Figure 8 Gene modules affected by the onset of lineage.

(a,b) Averaged gene expression of all gene modules, including those omitted from the main figure, was smoothened and plotted against the degree of lineage-specific priming. Data is shown for individual 1 (a) and 2 (b). (c) Comparison between gene modules from individual 1 and 2. For each module from individual 1, the overlap with each module from individual 2 is shown. Due to the higher number of cells analysed, gene modules from individual 2 split up into multiple modules from individual 1, while modules from individual 1 overlap only with a single module from individual 2. Only genes discovered in both individuals were included in this analysis.

Supplementary information

Supplementary Information

Supplementary Information (PDF 17138 kb)

Supplementary Table 1

Supplementary Information (XLSX 35 kb)

Supplementary Table 2

Supplementary Information (XLSX 466 kb)

Supplementary Table 3

Supplementary Information (XLSX 51 kb)

Supplementary Table 4

Supplementary Information (XLSX 1134 kb)

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Velten, L., Haas, S., Raffel, S. et al. Human haematopoietic stem cell lineage commitment is a continuous process. Nat Cell Biol 19, 271–281 (2017).

Download citation

Further reading


Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing