Lineage tracing on transcriptional landscapes links state to fate during differentiation

A challenge in biology is to associate molecular differences among progenitor cells with their capacity to generate mature cell types. Here, we use expressed DNA barcodes to clonally trace transcriptomes over time, applied to study fate determination in hematopoiesis. We identify states of primed fate potential, and locate them on a continuous transcriptional landscape. We identify two routes of monocyte differentiation that leave an imprint on mature cells. Clonal analysis also reveals fate biases into multiple lineages that depend on stable cellular properties hidden from single-cell RNA sequencing. Finally, we benchmark computational methods of dynamic inference from single-cell snapshots, showing that fate choice occurs earlier than is detected by state-of the-art algorithms, and that cells progress steadily through pseudotime with precise and consistent dynamics. Single-cell mRNA sequencing of fresh and cultured progenitors from mouse bone marrow. Cells were harvested, transduced with lineage barcodes, and then cultured or transplanted with single-cell RNA seq profiling across several time points. Sample stateFate_inVitro contains cells cultured in a broad cytokine cocktail. Sample stateFate_inVivo contains cells that were profiled before and after transplantation. Sample stateFate_cytokinePerturbation contains cells cultured in several different cytokine conditions. Sample GMP_MDP_culture contains MDPs and GMPs before after after culture. Raw sequencing data are available via SRA. All files are already demultiplexed. Read labels in the fastq files have format @[cell barcode]:[UMI]:[remainder]