Single-cell RNA sequencing (scRNA-seq) can characterize cell types and states through unsupervised clustering, but the ever increasing number of cells and batch effect impose computational challenges. We present DESC, an unsupervised deep embedding algorithm that clusters scRNA-seq data by iteratively optimizing a clustering objective function. Through iterative self-learning, DESC gradually removes batch effects, as long as technical differences across batches are smaller than true biological variations. As a soft clustering algorithm, cluster assignment probabilities from DESC are biologically interpretable and can reveal both discrete and pseudotemporal structure of cells. Comprehensive evaluations show that DESC offers a proper balance of clustering accuracy and stability, has a small footprint on memory, does not explicitly require batch information for batch effect removal, and can utilize GPU when available. As the scale of single-cell studies continues to grow, we believe DESC will offer a valuable tool for biomedical researchers to disentangle complex cellular heterogeneity. Increasingly large scRNA-seq datasets demand better and more scalable analysis tools. Here, the authors introduce a scalable unsupervised deep embedding algorithm that clusters scRNA-seq data by iteratively optimizing a clustering objective function and enables removal of batch effects.

In this issue of Cell, Uderhardt et al. employed intravital two-photon microscopy to examine tissue-resident macrophage responses to sterile cellular injuries of variable size. They observed that while multi-cell “macrolesions” are characteristically pro-inflammatory, resident macrophages can “cloak” single-cell microlesions to prevent excessive neutrophil recruitment and limit subsequent tissue damage.

The bone marrow microenvironment has a key role in regulating haematopoiesis, but its molecular complexity and response to stress are incompletely understood. Here we map the transcriptional landscape of mouse bone marrow vascular, perivascular and osteoblast cell populations at single-cell resolution, both at homeostasis and under conditions of stress-induced haematopoiesis. This analysis revealed previously unappreciated levels of cellular heterogeneity within the bone marrow niche and resolved cellular sources of pro-haematopoietic growth factors, chemokines and membrane-bound ligands. Our studies demonstrate a considerable transcriptional remodelling of niche elements under stress conditions, including an adipocytic skewing of perivascular cells. Among the stress-induced changes, we observed that vascular Notch delta-like ligands (encoded by Dll1 and Dll4) were downregulated. In the absence of vascular Dll4, haematopoietic stem cells prematurely induced a myeloid transcriptional program. These findings refine our understanding of the cellular architecture of the bone marrow niche, reveal a dynamic and heterogeneous molecular landscape that is highly sensitive to stress and illustrate the utility of single-cell transcriptomic data in evaluating the regulation of haematopoiesis by discrete niche populations. The transcriptional landscape of cell populations of the mouse bone marrow microenvironment, mapped at single-cell resolution, reveals cellular heterogeneity in this niche as well as substantial transcriptional remodelling under stress conditions.

Single cell analyses of mouse type I spiral ganglion neurons characterize three functionally distinct subtypes, revealing insights into auditory processing with implications for treating congenital deafness.

RNA sequencing of single cells in the mammalian trachea reveals a previously unknown airway cell that expresses genes involved in fluid and solute balance, and that might play a part in cystic fibrosis.

Single-cell analysis of melanoma tumor immune infiltrates reveals a separation between bystander cytotoxic T cells and a population that displays a continuous progression from a transitional toward a dysfunctional state that is associated with active proliferation and tumor reactivity.

The airways of the lung are the primary sites of disease in asthma and cystic fibrosis. Here we study the cellular composition and hierarchy of the mouse tracheal epithelium by single-cell RNA-sequencing (scRNA-seq) and in vivo lineage tracing. We identify a rare cell type, the Foxi1+ pulmonary ionocyte; functional variations in club cells based on their location; a distinct cell type in high turnover squamous epithelial structures that we term ‘hillocks’; and disease-relevant subsets of tuft and goblet cells. We developed ‘pulse-seq’, combining scRNA-seq and lineage tracing, to show that tuft, neuroendocrine and ionocyte cells are continually and directly replenished by basal progenitor cells. Ionocytes are the major source of transcripts of the cystic fibrosis transmembrane conductance regulator in both mouse (Cftr) and human (CFTR). Knockout of Foxi1 in mouse ionocytes causes loss of Cftr expression and disrupts airway fluid and mucus physiology, phenotypes that are characteristic of cystic fibrosis. By associating cell-type-specific expression programs with key disease genes, we establish a new cellular narrative for airways disease. Single-cell RNA sequencing analysis identifies cell types and lineages in airway epithelium, including the pulmonary ionocyte, a new cell type predominantly expressing the cystic fibrosis gene CFTR.

Single-cell transcriptomics and long-term lineage tracing outline aging-induced molecular identity changes of skin fibroblasts that lead them to acquire an adipogenic profile sensitive to whole-body metabolic changes.

Specification of a single precursor cell fate in zebrafish oogenesis is mediated by lateral inhibition driven by mechanical forces that lead to nuclear exclusion of a developmental transcriptional activator in neighboring cells.

Integrative analysis of single-cell transcriptomics and chromatin accessibility provides insights into regulatory features and dynamics in human hematopoiesis.

Welch et al. and Stuart et al. present novel techniques for the integration of single-cell RNA-seq datasets across multiple platforms, individuals, and species. They both extend these strategies to map cell types between RNA-seq datasets with epigenetic properties and in situ transcript profiling. The ability to transfer information between datasets and spatial methods will enable more comprehensive profiling and comparisons of cell populations in complex biological systems.

An integrative whole-organism single-cell analysis in Nematostella vectensis reveals cnidarian cell type complexity and provides insights into the mechanisms of animal cell-specific genomic regulation and evolution.

Live imaging of influenza A viral particles demonstrates that virion protein composition and morphology varies widely within the population produced by a single infected cell, suggesting that stochastic viral assembly may mediate phenotypic diversity and viral adaptation.

T cell development and selection are coordinated in the thymus by a specialized niche of diverse stromal populations1–3. Although much progress has been made over the years in identifying the functions of the different cell types of the thymic stromal compartment, there is no comprehensive characterization of their diversity and heterogeneity. Here we combined massively parallel single-cell RNA-sequencing4,5, spatial mapping, chromatin profiling and gene targeting to characterize de novo the entire stromal compartment of the mouse thymus. We identified dozens of cell states, with thymic epithelial cells (TECs) showing the highest degree of heterogeneity. Our analysis highlights four major medullary TEC (mTEC I–IV) populations, with distinct molecular functions, epigenetic landscapes and lineage regulators. Specifically, mTEC IV constitutes a new and highly divergent TEC lineage with molecular characteristics of the gut chemosensory epithelial tuft cells. Mice deficient in Pou2f3, a master regulator of tuft cells, have complete and specific depletion of mTEC IV cells, which results in increased levels of thymus-resident type-2 innate lymphoid cells. Overall, our study provides a comprehensive characterization of the thymic stroma and identifies a new tuft-like TEC population, which is critical for shaping the immune niche in the thymus. A comprehensive characterization of the thymic stroma identifies a tuft-cell-like thymic epithelial cell population that is critical for shaping the immune niche in the thymus.

Using single-cell RNA sequencing of both immune and non-immune cells in the developing lung, Cohen et al. map candidate cell-cell interactions during alveolar macrophage development. This revealed potential cross-talk between epithelial cells, ILC2s, basophils, and the developing macrophages, which were validated both in vitro and in vivo.

Here we present a compendium of single-cell transcriptomic data from the model organism Mus musculus that comprises more than 100,000 cells from 20 organs and tissues. These data represent a new resource for cell biology, reveal gene expression in poorly characterized cell populations and enable the direct and controlled comparison of gene expression in cell types that are shared between tissues, such as T lymphocytes and endothelial cells from different anatomical locations. Two distinct technical approaches were used for most organs: one approach, microfluidic droplet-based 3′-end counting, enabled the survey of thousands of cells at relatively low coverage, whereas the other, full-length transcript analysis based on fluorescence-activated cell sorting, enabled the characterization of cell types with high sensitivity and coverage. The cumulative data provide the foundation for an atlas of transcriptomic cell biology. A ‘mouse atlas’, comprising single-cell transcriptomic data from more than 100,000 cells from 20 organs and tissues, has been created as a resource for cell biology.

Direct lineage reprogramming involves the conversion of cellular identity. Single-cell technologies are useful for deconstructing the considerable heterogeneity that emerges during lineage conversion. However, lineage relationships are typically lost during cell processing, complicating trajectory reconstruction. Here we present ‘CellTagging’, a combinatorial cell-indexing methodology that enables parallel capture of clonal history and cell identity, in which sequential rounds of cell labelling enable the construction of multi-level lineage trees. CellTagging and longitudinal tracking of fibroblast to induced endoderm progenitor reprogramming reveals two distinct trajectories: one leading to successfully reprogrammed cells, and one leading to a ‘dead-end’ state, paths determined in the earliest stages of lineage conversion. We find that expression of a putative methyltransferase, Mettl7a1, is associated with the successful reprogramming trajectory; adding Mettl7a1 to the reprogramming cocktail increases the yield of induced endoderm progenitors. Together, these results demonstrate the utility of our lineage-tracing method for revealing the dynamics of direct reprogramming. Combinatorial tagging of single cells using expressed DNA barcodes, delivered by a lentiviral vector, is used to track individual cells and reconstruct their lineages and trajectories during cell fate reprogramming.

A combination of transcriptomics and mutational analyses in single cells from acute myeloid leukemia patients reveals the existence of distinct functional subsets and their associated drivers.

Using single-cell sequencing technologies, somatic mutations in mtDNA can be used as natural genetic barcodes to study cellular states and clonal dynamics.

Protein-level genetic barcodes enable single-cell high-dimensional phenotyping by mass cytometry in CRISPR screens.