Hematopoietic differentiation is characterized by a transient peak of entropy at a single cell level.

Hematopoietic differentiation has been metaphorically represented as linear trajectories with discrete steps from hematopoietic stem cells to mature cells. While the transcriptional state of cells at the beginning or at the end of these trajectories are well described from bulk analysis, what happens in the intermediate states has remained elusive until the use of single cell approaches. Applying Shannon entropy to measure cell-to-cell variability among cells at the same stage of differentiation, we observed a transient peak of gene expression variability in all the hematopoietic differentiation pathways. Strikingly, genes with the highest entropy variation in a given differentiation pathway matched genes known as pathway-specific, whereas genes with the highest expression variation were common to all pathways. Finally, we showed that the level of cell-to-cell variation is increased in the most immature compartment of hematopoiesis in myelodysplastic syndromes. These data suggest that differentiation could be better conceptualized as a dynamical stochastic process with a transient stage of cellular indetermination.