The human liver is an essential multifunctional organ, and liver diseases are rising with limited treatment options. However, the cellular composition of the liver remains poorly understood. Here, we performed single-cell RNA-sequencing of ~10,000 cells from normal liver tissue of 9 human donors to construct a human liver cell atlas. Our analysis revealed previously unknown sub-types among endothelial cells, Kupffer cells, and hepatocytes with transcriptome-wide zonation of some of these populations. We reveal heterogeneity of the EPCAM + population, which comprises hepatocyte-biased and cholangiocyte populations as well as a TROP2 int progenitor population with strong potential to form bipotent liver organoids. As proof-of-principle, we utilized our atlas to unravel phenotypic changes in hepatocellular carcinoma cells and in human hepatocytes and liver endothelial cells engrafted into a mouse liver. Our human liver cell atlas provides a powerful resource enabling the discovery of previously unknown cell types in the normal and diseased liver.

To understand stem cell differentiation along multiple lineages, it is necessary to resolve heterogeneous cellular states and the ancestral relationships between them. We developed a robotic miniaturized CEL-Seq2 implementation to carry out deep single-cell RNA-seq of ∼2,000 mouse hematopoietic progenitors enriched for lymphoid lineages, and used an improved clustering algorithm, RaceID3, to identify cell types. To resolve subtle transcriptome differences indicative of lineage biases, we developed FateID, an iterative supervised learning algorithm for the probabilistic quantification of cell fate bias in progenitor populations. Here we used FateID to delineate domains of fate bias and enable the derivation of high-resolution differentiation trajectories, thereby revealing a common progenitor population of B cells and plasmacytoid dendritic cells, which we validated by in vitro differentiation assays. We expect that FateID will improve understanding of the process of cell fate choice in complex multi-lineage differentiation systems.

Differentiation of multipotent cells is a complex process governed by interactions of thousands of genes subject to substantial expression fluctuations. Resolving cell state heterogeneity arising during this process requires quantification of gene expression within individual cells. However, computational methods linking this heterogeneity to biases towards distinct cell fates are not well established. Here, we perform deep single-cell transcriptome sequencing of ~2,000 bone-marrow derived mouse hematopoietic progenitors enriched for lymphoid lineages. To resolve subtle transcriptome priming indicative of distinct lineage biases, we developed FateID, an iterative supervised learning algorithm for the probabilistic quantification of cell fate bias. FateID delineates domains of fate bias within progenitor populations and permits the derivation of high-resolution differentiation trajectories, revealing a common progenitor population of B cells and plasmacytoid dendritic cells, which we validated by in vitro differentiation assays. We expect that FateID will enhance our understanding of the process of cell fate choice in complex multi-lineage differentiation systems. INTRODUCTIONRecent studies utilizing scRNA-seq 1-5 and single-cell lineage tracing techniques 6-8 , call into question the traditional view of hematopoietic differentiation as a sequence of binary fate choices giving rise to a succession of increasingly fate restricted progenitor types 9 . Evidence from these studies rather suggests early cell fate priming starting at the level of multipotent progenitors (MPP) or even within the HSC pool. Pronounced heterogeneity of common myeloid progenitors (CMP) was elucidated with high resolution 1 , and an early fate bias emerging in human short term HSCs was suggested in a more recent study 2 . However, heterogeneity of lymphoid progenitors has not been well investigated with single-cell resolution. Since lymphoid progenitor heterogeneity was previously found by flow cytometry 10,11 , utilizing distinct sets of cell surface markers in combination with differentiation assays, we here perform an scRNA-seq analysis to comprehensively elucidate heterogeneity across lymphoid progenitors purified from the bone-marrow of adult mice. Although a number of methods for lineage reconstruction have been developed [12][13][14][15] , these algorithms are not specifically designed to uncover subtle transcriptome changes and uniquely assign cells to individual branches without accounting for multi-lineage bias. Weak transcriptome modulations also remain undiscovered by state-of-the-art clustering methods, which partition cells into groups without accounting for the co-existence of fate bias towards multiple lineages within individual cells. To elucidate the process of cell fate emergence, we introduce FateID, a computational method for the quantification of fate bias, manifested by subtle lineage specific transcriptome modulations within a multipotent progenitor . CC-BY-NC-ND 4.0 International license not peer-reviewed) is the aut...