Single-cell RNA-seq quantifies biological heterogeneity across both discrete cell types and continuous cell transitions. Partition-based graph abstraction (PAGA) provides an interpretable graph-like map of the arising data manifold, based on estimating connectivity of manifold partitions ( https://github.com/theislab/paga ). PAGA maps preserve the global topology of data, allow analyzing data at different resolutions, and result in much higher computational efficiency of the typical exploratory data analysis workflow. We demonstrate the method by inferring structure-rich cell maps with consistent topology across four hematopoietic datasets, adult planaria and the zebrafish embryo and benchmark computational performance on one million neurons.

Single-cell RNA sequencing (scRNA-seq) has enabled researchers to study gene expression at a cellular resolution. However, noise due to amplification and dropout may obstruct analyses, so scalable denoising methods for increasingly large but sparse scRNA-seq data are needed. We propose a deep count autoencoder network (DCA) to denoise scRNA-seq datasets. DCA takes the count distribution, overdispersion and sparsity of the data into account using a negative binomial noise model with or without zero-inflation, and nonlinear gene-gene dependencies are captured. Our method scales linearly with the number of cells and can, therefore, be applied to datasets of millions of cells. We demonstrate that DCA denoising improves a diverse set of typical scRNA-seq data analyses using simulated and real datasets. DCA outperforms existing methods for data imputation in quality and speed, enhancing biological discovery.

Human lungs enable efficient gas exchange and form an interface with the environment, which depends on mucosal immunity for protection against infectious agents. Tightly controlled interactions between structural and immune cells are required to maintain lung homeostasis. Here, we use single-cell transcriptomics to chart the cellular landscape of upper and lower airways and lung parenchyma in healthy lungs, and lower airways in asthmatic lungs. We report location-dependent airway epithelial cell states and a novel subset of tissue-resident memory T cells. In the lower airways of patients with asthma, mucous cell hyperplasia is shown to stem from a novel mucous ciliated cell state, as well as goblet cell hyperplasia. We report the presence of pathogenic effector type 2 helper T cells (T_H2) in asthmatic lungs and find evidence for type 2 cytokines in maintaining the altered epithelial cell states. Unbiased analysis of cell–cell interactions identifies a shift from airway structural cell communication in healthy lungs to a T_H2-dominated interactome in asthmatic lungs.

Aging promotes lung function decline and susceptibility to chronic lung diseases, which are the third leading cause of death worldwide. Here, we use single cell transcriptomics and mass spectrometry-based proteomics to quantify changes in cellular activity states across 30 cell types and chart the lung proteome of young and old mice. We show that aging leads to increased transcriptional noise, indicating deregulated epigenetic control. We observe cell type-specific effects of aging, uncovering increased cholesterol biosynthesis in type-2 pneumocytes and lipofibroblasts and altered relative frequency of airway epithelial cells as hallmarks of lung aging. Proteomic profiling reveals extracellular matrix remodeling in old mice, including increased collagen IV and XVI and decreased Fraser syndrome complex proteins and collagen XIV. Computational integration of the aging proteome with the single cell transcriptomes predicts the cellular source of regulated proteins and creates an unbiased reference map of the aging lung.

Splicing factors such as BUD31 are identified in a synthetic-lethal screen with cells overexpressing the transcription factor MYC; oncogenic MYC leads to an increase in pre-mRNA synthesis, and spliceosome inhibition impairs the growth and tumorigenicity of MYC-dependent breast cancers, suggesting that spliceosome components may be potential therapeutic targets for MYC-driven cancers. The transcription factor MYC is frequently amplified or overexpressed in cancer and drives increased RNA and protein production. Here, Thomas Westbrook and colleagues identify the splicing factor BUD31 in a synthetic lethal screen with cells overexpressing MYC and show that other splicing factors are also required for cells to tolerate overexpressed MYC. Oncogenic MYC leads to an increase in pre-mRNA synthesis, and inhibition of the spliceosome impairs the growth and tumorigenicity of MYC-dependent breast cancer cells. Spliceosome components may therefore be potential therapeutic targets for MYC-driven cancers. MYC (also known as c-MYC) overexpression or hyperactivation is one of the most common drivers of human cancer. Despite intensive study, the MYC oncogene remains recalcitrant to therapeutic inhibition. MYC is a transcription factor, and many of its pro-tumorigenic functions have been attributed to its ability to regulate gene expression programs1,2,3. Notably, oncogenic MYC activation has also been shown to increase total RNA and protein production in many tissue and disease contexts4,5,6,7. While such increases in RNA and protein production may endow cancer cells with pro-tumour hallmarks, this increase in synthesis may also generate new or heightened burden on MYC-driven cancer cells to process these macromolecules properly8. Here we discover that the spliceosome is a new target of oncogenic stress in MYC-driven cancers. We identify BUD31 as a MYC-synthetic lethal gene in human mammary epithelial cells, and demonstrate that BUD31 is a component of the core spliceosome required for its assembly and catalytic activity. Core spliceosomal factors (such as SF3B1 and U2AF1) associated with BUD31 are also required to tolerate oncogenic MYC. Notably, MYC hyperactivation induces an increase in total precursor messenger RNA synthesis, suggesting an increased burden on the core spliceosome to process pre-mRNA. In contrast to normal cells, partial inhibition of the spliceosome in MYC-hyperactivated cells leads to global intron retention, widespread defects in pre-mRNA maturation, and deregulation of many essential cell processes. Notably, genetic or pharmacological inhibition of the spliceosome in vivo impairs survival, tumorigenicity and metastatic proclivity of MYC-dependent breast cancers. Collectively, these data suggest that oncogenic MYC confers a collateral stress on splicing, and that components of the spliceosome may be therapeutic entry points for aggressive MYC-driven cancers.

Activation of endoplasmic reticulum (ER) stress/the unfolded protein response (UPR) has been linked to cancer, but the molecular mechanisms are poorly understood and there is a paucity of reagents to translate this for cancer therapy. Here, we report that an IRE1α RNase-specific inhibitor, MKC8866, strongly inhibits prostate cancer (PCa) tumor growth as monotherapy in multiple preclinical models in mice and shows synergistic antitumor effects with current PCa drugs. Interestingly, global transcriptomic analysis reveal that IRE1α-XBP1s pathway activity is required for c-MYC signaling, one of the most highly activated oncogenic pathways in PCa. XBP1s is necessary for optimal c-MYC mRNA and protein expression, establishing, for the first time, a direct link between UPR and oncogene activation. In addition, an XBP1-specific gene expression signature is strongly associated with PCa prognosis. Our data establish IRE1α-XBP1s signaling as a central pathway in PCa and indicate that its targeting may offer novel treatment strategies.