The triumphs and limitations of computational methods for scRNA-seq

Single-cell and spatial transcriptomics: deciphering brain complexity in health and disease

In the past decade, single-cell technologies have proliferated and improved from their technically challenging beginnings to become common laboratory methods capable of determining the expression of thousands of genes in thousands of cells simultaneously. The field has progressed by taking the CNS as a primary research subject - the cellular complexity and multiplicity of neuronal cell types provide fertile ground for the increasing power of single-cell methods. Current single-cell RNA sequencing methods can quantify gene expression with sufficient accuracy to finely resolve even subtle differences between cell types and states, thus providing a great tool for studying the molecular and cellular repertoire of the CNS and its disorders. However, single-cell RNA sequencing requires the dissociation of tissue samples, which means that the interrelationships between cells are lost. Spatial transcriptomic methods bypass tissue dissociation and retain this spatial information, thereby allowing gene expression to be assessed across thousands of cells within the context of tissue structural organization. Here, we discuss how single-cell and spatially resolved transcriptomics have been contributing to unravelling the pathomechanisms underlying brain disorders. We focus on three areas where we feel these new technologies have provided particularly useful insights: selective neuronal vulnerability, neuroimmune dysfunction and cell-type-specific treatment response. We also discuss the limitations and future directions of single-cell and spatial RNA sequencing technologies.

Cellular Diversity in Human Subgenual Anterior Cingulate and Dorsolateral Prefrontal Cortex by Single-Nucleus RNA-Sequencing

Regional cellular heterogeneity is a fundamental feature of the human neocortex; however, details of this heterogeneity are still undefined. We used single-nucleus RNA-sequencing to examine cell-specific transcriptional features in the dorsolateral PFC (DLPFC) and the subgenual anterior cingulate cortex (sgACC), regions implicated in major psychiatric disorders. Droplet-based nuclei-capture and library preparation were performed on replicate samples from 8 male donors without history of psychiatric or neurologic disorder. Unsupervised clustering identified major neural cell classes. Subsequent iterative clustering of neurons further revealed 20 excitatory and 22 inhibitory subclasses. Inhibitory cells were consistently more abundant in the sgACC and excitatory neuron subclusters exhibited considerable variability across brain regions. Excitatory cell subclasses also exhibited greater within-class transcriptional differences between the two regions. We used these molecular definitions to determine which cell classes might be enriched in loci carrying a genetic signal in genome-wide association studies or for differentially expressed genes in mental illness. We found that the heritable signals of psychiatric disorders were enriched in neurons and that, while the gene expression changes detected in bulk-RNA-sequencing studies were dominated by glial cells, some alterations could be identified in specific classes of excitatory and inhibitory neurons. Intriguingly, only two excitatory cell classes exhibited concomitant region-specific enrichment for both genome-wide association study loci and transcriptional dysregulation. In sum, by detailing the molecular and cellular diversity of the DLPFC and sgACC, we were able to generate hypotheses on regional and cell-specific dysfunctions that may contribute to the development of mental illness.SIGNIFICANCE STATEMENT Dysfunction of the subgenual anterior cingulate cortex has been implicated in mood disorders, particularly major depressive disorder, and the dorsolateral PFC, a subsection of the PFC involved in executive functioning, has been implicated in schizophrenia. Understanding the cellular composition of these regions is critical to elucidating the neurobiology underlying psychiatric and neurologic disorders. We studied cell type diversity of the subgenual anterior cingulate cortex and dorsolateral PFC of humans with no neuropsychiatric illness using a clustering analysis of single-nuclei RNA-sequencing data. Defining the transcriptomic profile of cellular subpopulations in these cortical regions is a first step to demystifying the cellular and molecular pathways involved in psychiatric disorders.

Deciphering and advancing CAR T-cell therapy with single-cell sequencing technologies

Chimeric antigen receptor (CAR) T-cell therapy has made remarkable progress in cancer immunotherapy, but several challenges with unclear mechanisms hinder its wide clinical application. Single-cell sequencing technologies, with the powerful unbiased analysis of cellular heterogeneity and molecular patterns at unprecedented resolution, have greatly advanced our understanding of immunology and oncology. In this review, we summarize the recent applications of single-cell sequencing technologies in CAR T-cell therapy, including the biological characteristics, the latest mechanisms of clinical response and adverse events, promising strategies that contribute to the development of CAR T-cell therapy and CAR target selection. Generally, we propose a multi-omics research mode to guide potential future research on CAR T-cell therapy.

Acoel single-cell atlas reveals expression dynamics and heterogeneity of adult pluripotent stem cells

Adult pluripotent stem cell (aPSC) populations underlie whole-body regeneration in many distantly-related animal lineages, but how the underlying cellular and molecular mechanisms compare across species is unknown. Here, we apply single-cell RNA sequencing to profile transcriptional cell states of the acoel worm Hofstenia miamia during postembryonic development and regeneration. We identify cell types shared across stages and their associated gene expression dynamics during regeneration. Functional studies confirm that the aPSCs, also known as neoblasts, are the source of differentiated cells and reveal transcription factors needed for differentiation. Subclustering of neoblasts recovers transcriptionally distinct subpopulations, the majority of which are likely specialized to differentiated lineages. One neoblast subset, showing enriched expression of the histone variant H3.3, appears to lack specialization. Altogether, the cell states identified in this study facilitate comparisons to other species and enable future studies of stem cell fate potentials.

scRNASequest: an ecosystem of scRNA-seq analysis, visualization, and publishing

BACKGROUND

Single-cell RNA sequencing is a state-of-the-art technology to understand gene expression in complex tissues. With the growing amount of data being generated, the standardization and automation of data analysis are critical to generating hypotheses and discovering biological insights.

RESULTS

Here, we present scRNASequest, a semi-automated single-cell RNA-seq (scRNA-seq) data analysis workflow which allows (1) preprocessing from raw UMI count data, (2) harmonization by one or multiple methods, (3) reference-dataset-based cell type label transfer and embedding projection, (4) multi-sample, multi-condition single-cell level differential gene expression analysis, and (5) seamless integration with cellxgene VIP for visualization and with CellDepot for data hosting and sharing by generating compatible h5ad files.

CONCLUSIONS

We developed scRNASequest, an end-to-end pipeline for single-cell RNA-seq data analysis, visualization, and publishing. The source code under MIT open-source license is provided at https://github.com/interactivereport/scRNASequest . We also prepared a bookdown tutorial for the installation and detailed usage of the pipeline: https://interactivereport.github.io/scRNAsequest/tutorial/docs/ . Users have the option to run it on a local computer with a Linux/Unix system including MacOS, or interact with SGE/Slurm schedulers on high-performance computing (HPC) clusters.

Isolation of human cutaneous immune cells for single-cell RNA sequencing

Single-cell RNA sequencing (scRNA-seq) allows for high-resolution analysis of transcriptionally dysregulated cell subpopulations in inflammatory diseases. However, it can be challenging to properly isolate viable immune cells from human skin for scRNA-seq due to its barrier properties. Here, we present a protocol to isolate high-viability human cutaneous immune cells. We describe steps for obtaining and enzymatically dissociating a skin biopsy specimen and isolating immune cells using flow cytometry. We then provide an overview of downstream computational techniques to analyze sequencing data. For complete details on the use and execution of this protocol, please refer to Cook et al. (2022)1 and Liu et al. (2022).2.

Comparison of transformations for single-cell RNA-seq data

The count table, a numeric matrix of genes × cells, is the basic input data structure in the analysis of single-cell RNA-sequencing data. A common preprocessing step is to adjust the counts for variable sampling efficiency and to transform them so that the variance is similar across the dynamic range. These steps are intended to make subsequent application of generic statistical methods more palatable. Here, we describe four transformation approaches based on the delta method, model residuals, inferred latent expression state and factor analysis. We compare their strengths and weaknesses and find that the latter three have appealing theoretical properties; however, in benchmarks using simulated and real-world data, it turns out that a rather simple approach, namely, the logarithm with a pseudo-count followed by principal-component analysis, performs as well or better than the more sophisticated alternatives. This result highlights limitations of current theoretical analysis as assessed by bottom-line performance benchmarks.

SEQUIN is an R/Shiny framework for rapid and reproducible analysis of RNA-seq data

SEQUIN is a web-based application (app) that allows fast and intuitive analysis of RNA sequencing data derived for model organisms, tissues, and single cells. Integrated app functions enable uploading datasets, quality control, gene set enrichment, data visualization, and differential gene expression analysis. We also developed the iPSC Profiler, a practical gene module scoring tool that helps measure and compare pluripotent and differentiated cell types. Benchmarking to other commercial and non-commercial products underscored several advantages of SEQUIN. Freely available to the public, SEQUIN empowers scientists using interdisciplinary methods to investigate and present transcriptome data firsthand with state-of-the-art statistical methods. Hence, SEQUIN helps democratize and increase the throughput of interrogating biological questions using next-generation sequencing data with single-cell resolution.

Single-cell profiling to explore pancreatic cancer heterogeneity, plasticity and response to therapy

Pancreatic ductal adenocarcinoma (PDAC) is a highly lethal cancer entity characterized by a heterogeneous genetic landscape and an immunosuppressive tumor microenvironment. Recent advances in high-resolution single-cell sequencing and spatial transcriptomics technologies have enabled an in-depth characterization of both malignant and host cell types and increased our understanding of the heterogeneity and plasticity of PDAC in the steady state and under therapeutic perturbation. In this Review we outline single-cell analyses in PDAC, discuss their implications on our understanding of the disease and present future perspectives of multimodal approaches to elucidate its biology and response to therapy at the single-cell level.

Ensemble learning-based gene signature and risk model for predicting prognosis of triple-negative breast cancer

Although medical science has been fully developed, due to the high heterogeneity of triple-negative breast cancer (TNBC), it is still difficult to use reasonable and precise treatment. In this study, based on local optimization-feature screening and genomics screening strategy, we screened 25 feature genes. In multiple machine learning algorithms, feature genes have excellent discriminative diagnostic performance among samples composed of multiple large datasets. After screening at the single-cell level, we identified genes expressed substantially in myeloid cells (MCGs) that have a potential association with TNBC. Based on MCGs, we distinguished two types of TNBC patients who showed considerable differences in survival status and immune-related characteristics. Immune-related gene risk scores (IRGRS) were established, and their validity was verified using validation cohorts. A total of 25 feature genes were obtained, among which CXCL9, CXCL10, CCL7, SPHK1, and TREM1 were identified as the result after single-cell level analysis and screening. According to these entries, the cohort was divided into MCA and MCB subtypes, and the two subtypes had significant differences in survival status and tumor-immune microenvironment. After Lasso-Cox screening, IDO1, GNLY, IRF1, CTLA4, and CXCR6 were selected for constructing IRGRS. There were significant differences in drug sensitivity and immunotherapy sensitivity among high-IRGRS and low-IRGRS groups. We revealed the dynamic relationship between TNBC and TIME, identified a potential biomarker called Granulysin (GNLY) related to immunity, and developed a multi-process machine learning package called "MPMLearning 1.0" in Python.

Single-cell sequencing reveals that endothelial cells, EndMT cells and mural cells contribute to the pathogenesis of cavernous malformations

Cavernous malformations (CMs) invading the central nervous system occur in ~0.16-0.4% of the general population, often resulting in hemorrhages and focal neurological deficits. Further understanding of disease mechanisms and therapeutic strategies requires a deeper knowledge of CMs in humans. Herein, we performed single-cell RNA sequencing (scRNA-seq) analysis on unselected viable cells from twelve human CM samples and three control samples. A total of 112,670 high-quality cells were clustered into 11 major cell types, which shared a number of common features in CMs harboring different genetic mutations. A new EC subpopulation marked with PLVAP was uniquely identified in lesions. The cellular ligand‒receptor network revealed that the PLVAP-positive EC subcluster was the strongest contributor to the ANGPT and VEGF signaling pathways in all cell types. The PI3K/AKT/mTOR pathway was strongly activated in the PLVAP-positive subcluster even in non-PIK3CA mutation carriers. Moreover, endothelial-to-mesenchymal transition (EndMT) cells were identified for the first time in CMs at the single-cell level, which was accompanied by strong immune activation. The transcription factor SPI1 was predicted to be a novel key driver of EndMT, which was confirmed by in vitro and in vivo studies. A specific fibroblast-like phenotype was more prevalent in lesion smooth muscle cells, hinting at the role of vessel reconstructions and repairs in CMs, and we also confirmed that TWIST1 could induce SMC phenotypic switching in vitro and in vivo. Our results provide novel insights into the pathomechanism decryption and further precise therapy of CMs.

Single-cell RNA sequencing to identify cellular heterogeneity and targets in cardiovascular diseases: from bench to bedside

The mechanisms of cardiovascular diseases (CVDs) remain incompletely elucidated. Single-cell RNA sequencing (scRNA-seq) has enabled the profiling of single-cell transcriptomes at unprecedented resolution and throughput, which is critical for deciphering cardiovascular cellular heterogeneity and underlying disease mechanisms, thereby facilitating the development of therapeutic strategies. In this review, we summarize cellular heterogeneity in cardiovascular homeostasis and diseases as well as the discovery of potential disease targets based on scRNA-seq, and yield new insights into the promise of scRNA-seq technology in precision medicine and clinical application.

Molecular recording: transcriptional data collection into the genome

Advances in regenerative medicine depend upon understanding the complex transcriptional choreography that guides cellular development. Transcriptional molecular recorders, tools that record different transcriptional events into the genome of cells, hold promise to elucidate both the intensity and timing of transcriptional activity at single-cell resolution without requiring destructive multitime point assays. These technologies are dependent on DNA writers, which translate transcriptional signals into stable genomic mutations that encode the duration, intensity, and order of transcriptional events. In this review, we highlight recent progress toward more informative and multiplexable transcriptional recording through the use of three different types of DNA writing - recombineering, Cas1-Cas2 acquisition, and prime editing - and the architecture of the genomic data generated.

Multiomic Spatial Mapping of Myocardial Infarction and Implications for Personalized Therapy

Ischemic heart disease including myocardial infarction is still the leading cause of death worldwide. Although the survival early after myocardial infarction has been significantly improved by the introduction of percutaneous coronary intervention, long-term morbidity and mortality remain high. The elevated long-term mortality is mainly driven by cardiac remodeling processes triggering ischemic heart failure and electric instability. Despite the new developments in pharmaco-therapy of heart failure, we still lack targeted therapies for cardiac remodeling and fibrosis. Single-cell and genomic technologies allow us to map the human heart at unprecedented resolution and allow to gain insights into cellular and molecular heterogeneity. However, these technologies rely on digested tissue and isolated cells or nuclei and thus lack spatial information. Spatial information is critical to understand tissue homeostasis and disease and can be utilized to identify disease-driving cell populations and mechanisms including cellular cross-talk. Here, we discuss recent advances in single-cell and spatial genomic technologies that give insights into cellular and molecular mechanisms of cardiac remodeling after injury and can be utilized to identify novel therapeutic targets and pave the way toward new therapies in heart failure.

scMINER: a mutual information-based framework for identifying hidden drivers from single-cell omics data

The sparse nature of single-cell omics data makes it challenging to dissect the wiring and rewiring of the transcriptional and signaling drivers that regulate cellular states. Many of the drivers, referred to as "hidden drivers", are difficult to identify via conventional expression analysis due to low expression and inconsistency between RNA and protein activity caused by post-translational and other modifications. To address this issue, we developed scMINER, a mutual information (MI)-based computational framework for unsupervised clustering analysis and cell-type specific inference of intracellular networks, hidden drivers and network rewiring from single-cell RNA-seq data. We designed scMINER to capture nonlinear cell-cell and gene-gene relationships and infer driver activities. Systematic benchmarking showed that scMINER outperforms popular single-cell clustering algorithms, especially in distinguishing similar cell types. With respect to network inference, scMINER does not rely on the binding motifs which are available for a limited set of transcription factors, therefore scMINER can provide quantitative activity assessment for more than 6,000 transcription and signaling drivers from a scRNA-seq experiment. As demonstrations, we used scMINER to expose hidden transcription and signaling drivers and dissect their regulon rewiring in immune cell heterogeneity, lineage differentiation, and tissue specification. Overall, activity-based scMINER is a widely applicable, highly accurate, reproducible and scalable method for inferring cellular transcriptional and signaling networks in each cell state from scRNA-seq data. The scMINER software is publicly accessible via: https://github.com/jyyulab/scMINER.

Single-cell methods in myeloproliferative neoplasms: old questions, new technologies

Myeloproliferative neoplasms (MPN) are a group of clonal stem cell-derived hematopoietic malignancies driven by aberrant Janus kinase-signal transducer and activator of transcription proteins (JAK/STAT) signaling. Although these are genetically simple diseases, MPNs are phenotypically heterogeneous, reflecting underlying intratumoral heterogeneity driven by the interplay of genetic and nongenetic factors. Their evolution is determined by factors that enable certain cellular subsets to outcompete others. Therefore, techniques that resolve cellular heterogeneity at the single-cell level are ideally placed to provide new insights into MPN biology. With these insights comes the potential to uncover new approaches to predict the clinical course and treat these cancers, ultimately improving outcomes for patients. MPNs present a particularly tractable model of cancer evolution, because most patients present in an early disease phase and only a small proportion progress to aggressive disease. Therefore, it is not surprising that many groundbreaking technological advances in single-cell omics have been pioneered by their application in MPNs. In this review article, we explore how single-cell approaches have provided transformative insights into MPN disease biology, which are broadly applicable across human cancers, and discuss how these studies might be swiftly translated into clinical pathways and may eventually underpin precision medicine.

Understanding the interplay of membrane trafficking, cell surface mechanics, and stem cell differentiation

Stem cells can generate a diversity of cell types during development, regeneration and adult tissue homeostasis. Differentiation changes not only the cell fate in terms of gene expression but also the physical properties and functions of cells, e.g. the secretory activity, cell shape, or mechanics. Conversely, these activities and properties can also regulate differentiation itself. Membrane trafficking is known to modulate signal transduction and thus has the potential to control stem cell differentiation. On the other hand, membrane trafficking, particularly from and to the plasma membrane, depends on the mechanical properties of the cell surface such as tension within the plasma membrane or the cortex. Indeed, recent findings demonstrate that cell surface mechanics can also control cell fate. Here, we review the bidirectional relationships between these three fundamental cellular functions, i.e. membrane trafficking, cell surface mechanics, and stem cell differentiation. Furthermore, we discuss commonly used methods in each field and how combining them with new tools will enhance our understanding of their interplay. Understanding how membrane trafficking and cell surface mechanics can guide stem cell fate holds great potential as these concepts could be exploited for directed differentiation of stem cells for the fields of tissue engineering and regenerative medicine.

Decoding anterior-posterior axis emergence among mouse, monkey, and human embryos

Anterior-posterior axis formation regulated by the distal visceral endoderm (DVE) and anterior visceral endoderm (AVE) is essential for peri-implantation embryogenesis. However, the principles of the origin and specialization of DVE and AVE remain elusive. Here, with single-cell transcriptome analysis and pseudotime prediction, we show that DVE and AVE independently originate from the specialized primary endoderm in mouse blastocysts. Along distinct developmental paths, these two lineages, respectively, undergo four representative states with stage-specific transcriptional patterns around implantation. Further comparative analysis shows that AVE, but not DVE, is detected in human and non-human primate embryos, defining differences in polarity formation across species. Moreover, stem cell-assembled human blastoids lack DVE or AVE precursors, implying that additional induction of stem cells with DVE/AVE potential could promote the current embryo-like models and their post-implantation growth. Our work provides insight into understanding of embryonic polarity formation and early mammalian development.

Macrophages and microglia in glioblastoma: heterogeneity, plasticity, and therapy

Glioblastoma (GBM) is the most aggressive tumor in the central nervous system and contains a highly immunosuppressive tumor microenvironment (TME). Tumor-associated macrophages and microglia (TAMs) are a dominant population of immune cells in the GBM TME that contribute to most GBM hallmarks, including immunosuppression. The understanding of TAMs in GBM has been limited by the lack of powerful tools to characterize them. However, recent progress on single-cell technologies offers an opportunity to precisely characterize TAMs at the single-cell level and identify new TAM subpopulations with specific tumor-modulatory functions in GBM. In this Review, we discuss TAM heterogeneity and plasticity in the TME and summarize current TAM-targeted therapeutic potential in GBM. We anticipate that the use of single-cell technologies followed by functional studies will accelerate the development of novel and effective TAM-targeted therapeutics for GBM patients.

Clinically relevant fusion oncogenes: detection and practical implications

Mechanistically, chimeric genes result from DNA rearrangements and include parts of preexisting normal genes combined at the genomic junction site. Some rearranged genes encode pathological proteins with altered molecular functions. Those which can aberrantly promote carcinogenesis are called fusion oncogenes. Their formation is not a rare event in human cancers, and many of them were documented in numerous study reports and in specific databases. They may have various molecular peculiarities like increased stability of an oncogenic part, self-activation of tyrosine kinase receptor moiety, and altered transcriptional regulation activities. Currently, tens of low molecular mass inhibitors are approved in cancers as the drugs targeting receptor tyrosine kinase (RTK) oncogenic fusion proteins, that is, including ALK, ABL, EGFR, FGFR1-3, NTRK1-3, MET, RET, ROS1 moieties. Therein, the presence of the respective RTK fusion in the cancer genome is the diagnostic biomarker for drug prescription. However, identification of such fusion oncogenes is challenging as the breakpoint may arise in multiple sites within the gene, and the exact fusion partner is generally unknown. There is no gold standard method for RTK fusion detection, and many alternative experimental techniques are employed nowadays to solve this issue. Among them, RNA-seq-based methods offer an advantage of unbiased high-throughput analysis of only transcribed RTK fusion genes, and of simultaneous finding both fusion partners in a single RNA-seq read. Here we focus on current knowledge of biology and clinical aspects of RTK fusion genes, related databases, and laboratory detection methods.

Using "Galaxy-rCASC": A Public Galaxy Instance for Single-Cell RNA-Seq Data Analysis

rCASC is a modular workflow providing an integrated environment for single-cell RNA-seq (scRNA-Seq) data analysis exploiting Docker containers to achieve functional and computational reproducibility. It was initially developed as an R package usable also through a Java GUI. However, the Java frontend cannot be employed when running rCASC on a remote server, a typical setup due to the significant computational resources commonly needed to analyze scRNA-Seq data.To allow the use of rCASC through a graphical user interface on the client side and to harness the many advantages provided by the Galaxy platform, we have made rCASC available as a Galaxy set of tools, also providing a dedicated public instance of Galaxy named "Galaxy-rCASC." To integrate rCASC into Galaxy, all its functions, originally implemented as a set of Docker containers to maximize reproducibility, have been extensively reworked to become independent from the R package functions that launch them in the original implementation. Furthermore, suitable Galaxy wrappers have been developed for most functions of rCASC. We provide a detailed reference document to the use of Galaxy-rCASC with insights and explanations on the platform functionalities, parameters, and output while guiding the reader through the typical rCASC analysis workflow of a scRNA-Seq dataset.

Spatial-ID: a cell typing method for spatially resolved transcriptomics via transfer learning and spatial embedding

Spatially resolved transcriptomics provides the opportunity to investigate the gene expression profiles and the spatial context of cells in naive state, but at low transcript detection sensitivity or with limited gene throughput. Comprehensive annotating of cell types in spatially resolved transcriptomics to understand biological processes at the single cell level remains challenging. Here we propose Spatial-ID, a supervision-based cell typing method, that combines the existing knowledge of reference single-cell RNA-seq data and the spatial information of spatially resolved transcriptomics data. We present a series of benchmarking analyses on publicly available spatially resolved transcriptomics datasets, that demonstrate the superiority of Spatial-ID compared with state-of-the-art methods. Besides, we apply Spatial-ID on a self-collected mouse brain hemisphere dataset measured by Stereo-seq, that shows the scalability of Spatial-ID to three-dimensional large field tissues with subcellular spatial resolution.

Hypergraph geometry reflects higher-order dynamics in protein interaction networks

Protein interactions form a complex dynamic molecular system that shapes cell phenotype and function; in this regard, network analysis is a powerful tool for studying the dynamics of cellular processes. Current models of protein interaction networks are limited in that the standard graph model can only represent pairwise relationships. Higher-order interactions are well-characterized in biology, including protein complex formation and feedback or feedforward loops. These higher-order relationships are better represented by a hypergraph as a generalized network model. Here, we present an approach to analyzing dynamic gene expression data using a hypergraph model and quantify network heterogeneity via Forman-Ricci curvature. We observe, on a global level, increased network curvature in pluripotent stem cells and cancer cells. Further, we use local curvature to conduct pathway analysis in a melanoma dataset, finding increased curvature in several oncogenic pathways and decreased curvature in tumor suppressor pathways. We compare this approach to a graph-based model and a differential gene expression approach.

DeepVelo: Single-cell transcriptomic deep velocity field learning with neural ordinary differential equations

Recent advances in single-cell sequencing technologies have provided unprecedented opportunities to measure the gene expression profile and RNA velocity of individual cells. However, modeling transcriptional dynamics is computationally challenging because of the high-dimensional, sparse nature of the single-cell gene expression measurements and the nonlinear regulatory relationships. Here, we present DeepVelo, a neural network-based ordinary differential equation that can model complex transcriptome dynamics by describing continuous-time gene expression changes within individual cells. We apply DeepVelo to public datasets from different sequencing platforms to (i) formulate transcriptome dynamics on different time scales, (ii) measure the instability of cell states, and (iii) identify developmental driver genes via perturbation analysis. Benchmarking against the state-of-the-art methods shows that DeepVelo can learn a more accurate representation of the velocity field. Furthermore, our perturbation studies reveal that single-cell dynamical systems could exhibit chaotic properties. In summary, DeepVelo allows data-driven discoveries of differential equations that delineate single-cell transcriptome dynamics.

A high-efficiency differential expression method for cancer heterogeneity using large-scale single-cell RNA-sequencing data

Colorectal cancer is a highly heterogeneous disease. Tumor heterogeneity limits the efficacy of cancer treatment. Single-cell RNA-sequencing technology (scRNA-seq) is a powerful tool for studying cancer heterogeneity at cellular resolution. The sparsity, heterogeneous diversity, and fast-growing scale of scRNA-seq data pose challenges to the flexibility, accuracy, and computing efficiency of the differential expression (DE) methods. We proposed HEART (high-efficiency and robust test), a statistical combination test that can detect DE genes with various sources of differences beyond mean expression changes. To validate the performance of HEART, we compared HEART and the other six popular DE methods on various simulation datasets with different settings by two simulation data generation mechanisms. HEART had high accuracy (F1 score >0.75) and brilliant computational efficiency (less than 2 min) on multiple simulation datasets in various experimental settings. HEART performed well on DE genes detection for the PBMC68K dataset quantified by UMI counts and the human brain single-cell dataset quantified by read counts (F1 score = 0.79, 0.65). By applying HEART to the single-cell dataset of a colorectal cancer patient, we found several potential blood-based biomarkers (CTTN, S100A4, S100A6, UBA52, FAU, and VIM) associated with colorectal cancer metastasis and validated them on additional spatial transcriptomic data of other colorectal cancer patients.

Single-cell RNA-sequencing data analysis reveals a highly correlated triphasic transcriptional response to SARS-CoV-2 infection

Single-cell RNA sequencing (scRNA-seq) is currently one of the most powerful techniques available to study the transcriptional response of thousands of cells to an external perturbation. Here, we perform a pseudotime analysis of SARS-CoV-2 infection using publicly available scRNA-seq data from human bronchial epithelial cells and colon and ileum organoids. Our results reveal that, for most genes, the transcriptional response to SARS-CoV-2 infection follows a non-linear pattern characterized by an initial and a final down-regulatory phase separated by an intermediate up-regulatory stage. A correlation analysis of transcriptional profiles suggests a common mechanism regulating the mRNA levels of most genes. Interestingly, genes encoded in the mitochondria or involved in translation exhibited distinct pseudotime profiles. To explain our results, we propose a simple model where nuclear export inhibition of nsp1-sensitive transcripts will be sufficient to explain the transcriptional shutdown of SARS-CoV-2 infected cells.

UniTVelo: temporally unified RNA velocity reinforces single-cell trajectory inference

The recent breakthrough of single-cell RNA velocity methods brings attractive promises to reveal directed trajectory on cell differentiation, states transition and response to perturbations. However, the existing RNA velocity methods are often found to return erroneous results, partly due to model violation or lack of temporal regularization. Here, we present UniTVelo, a statistical framework of RNA velocity that models the dynamics of spliced and unspliced RNAs via flexible transcription activities. Uniquely, it also supports the inference of a unified latent time across the transcriptome. With ten datasets, we demonstrate that UniTVelo returns the expected trajectory in different biological systems, including hematopoietic differentiation and those even with weak kinetics or complex branches.

In situ tools for chromatin structural epigenomics

Technological progress over the past 15 years has fueled an explosion in genome-wide chromatin profiling tools that take advantage of low-cost short-read sequencing technologies to map particular chromatin features. Here, we survey the recent development of epigenomic tools that provide precise positions of chromatin proteins genome-wide in intact cells or nuclei. Some profiling tools are based on tethering Micrococcal Nuclease to chromatin proteins of interest in situ, whereas others similarly tether Tn5 transposase to integrate DNA sequencing adapters (tagmentation) and so eliminate the need for library preparation. These in situ cleavage and tagmentation tools have gained in popularity over the past few years, with many protocol enhancements and adaptations for single-cell and spatial chromatin profiling. The application of experimental and computational tools to address problems in gene regulation, eukaryotic development, and human disease are helping to define the emerging field of chromatin structural epigenomics.

The performance of deep generative models for learning joint embeddings of single-cell multi-omics data

Recent extensions of single-cell studies to multiple data modalities raise new questions regarding experimental design. For example, the challenge of sparsity in single-omics data might be partly resolved by compensating for missing information across modalities. In particular, deep learning approaches, such as deep generative models (DGMs), can potentially uncover complex patterns via a joint embedding. Yet, this also raises the question of sample size requirements for identifying such patterns from single-cell multi-omics data. Here, we empirically examine the quality of DGM-based integrations for varying sample sizes. We first review the existing literature and give a short overview of deep learning methods for multi-omics integration. Next, we consider eight popular tools in more detail and examine their robustness to different cell numbers, covering two of the most common multi-omics types currently favored. Specifically, we use data featuring simultaneous gene expression measurements at the RNA level and protein abundance measurements for cell surface proteins (CITE-seq), as well as data where chromatin accessibility and RNA expression are measured in thousands of cells (10x Multiome). We examine the ability of the methods to learn joint embeddings based on biological and technical metrics. Finally, we provide recommendations for the design of multi-omics experiments and discuss potential future developments.

Single-cell and single-nuclei RNA sequencing as powerful tools to decipher cellular heterogeneity and dysregulation in neurodegenerative diseases

Neurodegenerative diseases affect millions of people worldwide and there are currently no cures. Two types of common neurodegenerative diseases are Alzheimer's (AD) and Parkinson's disease (PD). Single-cell and single-nuclei RNA sequencing (scRNA-seq and snRNA-seq) have become powerful tools to elucidate the inherent complexity and dynamics of the central nervous system at cellular resolution. This technology has allowed the identification of cell types and states, providing new insights into cellular susceptibilities and molecular mechanisms underlying neurodegenerative conditions. Exciting research using high throughput scRNA-seq and snRNA-seq technologies to study AD and PD is emerging. Herein we review the recent progress in understanding these neurodegenerative diseases using these state-of-the-art technologies. We discuss the fundamental principles and implications of single-cell sequencing of the human brain. Moreover, we review some examples of the computational and analytical tools required to interpret the extensive amount of data generated from these assays. We conclude by highlighting challenges and limitations in the application of these technologies in the study of AD and PD.

Recent advances in the field of single-cell proteomics

The field of single-cell omics is rapidly progressing. Although DNA and RNA sequencing-based methods have dominated the field to date, global proteome profiling has also entered the main stage. Single-cell proteomics was facilitated by advancements in different aspects of mass spectrometry (MS)-based proteomics, such as instrument design, sample preparation, chromatography and ion mobility. Single-cell proteomics by mass spectrometry (scp-MS) has moved beyond being a mere technical development, and is now able to deliver actual biological application and has been successfully applied to characterize different cell states. Here, we review some key developments of scp-MS, provide a background to the field, discuss the various available methods and foresee possible future directions.

Single-Cell Immunobiology of the Maternal-Fetal Interface

Pregnancy success requires constant dialogue between the mother and developing conceptus. Such crosstalk is facilitated through complex interactions between maternal and fetal cells at distinct tissue sites, collectively termed the "maternal-fetal interface." The emergence of single-cell technologies has enabled a deeper understanding of the unique processes taking place at the maternal-fetal interface as well as the discovery of novel pathways and immune and nonimmune cell types. Single-cell approaches have also been applied to decipher the cellular dynamics throughout pregnancy, in parturition, and in obstetrical syndromes such as recurrent spontaneous abortion, preeclampsia, and preterm labor. Furthermore, single-cell technologies have been used during the recent COVID-19 pandemic to evaluate placental viral cell entry and the impact of SARS-CoV-2 infection on maternal and fetal immunity. In this brief review, we summarize the current knowledge of cellular immunobiology in pregnancy and its complications that has been generated through single-cell investigations of the maternal-fetal interface.

Single Cell Transcriptomics to Understand HSC Heterogeneity and Its Evolution upon Aging

Single-cell transcriptomic technologies enable the uncovering and characterization of cellular heterogeneity and pave the way for studies aiming at understanding the origin and consequences of it. The hematopoietic system is in essence a very well adapted model system to benefit from this technological advance because it is characterized by different cellular states. Each cellular state, and its interconnection, may be defined by a specific location in the global transcriptional landscape sustained by a complex regulatory network. This transcriptomic signature is not fixed and evolved over time to give rise to less efficient hematopoietic stem cells (HSC), leading to a well-documented hematopoietic aging. Here, we review the advance of single-cell transcriptomic approaches for the understanding of HSC heterogeneity to grasp HSC deregulations upon aging. We also discuss the new bioinformatics tools developed for the analysis of the resulting large and complex datasets. Finally, since hematopoiesis is driven by fine-tuned and complex networks that must be interconnected to each other, we highlight how mathematical modeling is beneficial for doing such interconnection between multilayered information and to predict how HSC behave while aging.

A pulmonologist's guide to perform and analyse cross-species single lung cell transcriptomics

Single-cell ribonucleic acid sequencing is becoming widely employed to study biological processes at a novel resolution depth. The ability to analyse transcriptomes of multiple heterogeneous cell types in parallel is especially valuable for cell-focused lung research where a variety of resident and recruited cells are essential for maintaining organ functionality. We compared the single-cell transcriptomes from publicly available and unpublished datasets of the lungs in six different species: human (Homo sapiens), African green monkey (Chlorocebus sabaeus), pig (Sus domesticus), hamster (Mesocricetus auratus), rat (Rattus norvegicus) and mouse (Mus musculus) by employing RNA velocity and intercellular communication based on ligand-receptor co-expression, among other techniques. Specifically, we demonstrated a workflow for interspecies data integration, applied a single unified gene nomenclature, performed cell-specific clustering and identified marker genes for each species. Overall, integrative approaches combining newly sequenced as well as publicly available datasets could help identify species-specific transcriptomic signatures in both healthy and diseased lung tissue and select appropriate models for future respiratory research.

Machine learning for cell type classification from single nucleus RNA sequencing data

With the advent of single cell/nucleus RNA sequencing (sc/snRNA-seq), the field of cell phenotyping is now a data-driven exercise providing statistical evidence to support cell type/state categorization. However, the task of classifying cells into specific, well-defined categories with the empirical data provided by sc/snRNA-seq remains nontrivial due to the difficulty in determining specific differences between related cell types with close transcriptional similarities, resulting in challenges with matching cell types identified in separate experiments. To investigate possible approaches to overcome these obstacles, we explored the use of supervised machine learning methods-logistic regression, support vector machines, random forests, neural networks, and light gradient boosting machine (LightGBM)-as approaches to classify cell types using snRNA-seq datasets from human brain middle temporal gyrus (MTG) and human kidney. Classification accuracy was evaluated using an F-beta score weighted in favor of precision to account for technical artifacts of gene expression dropout. We examined the impact of hyperparameter optimization and feature selection methods on F-beta score performance. We found that the best performing model for granular cell type classification in both datasets is a multinomial logistic regression classifier and that an effective feature selection step was the most influential factor in optimizing the performance of the machine learning pipelines.

Reconstructing data-driven governing equations for cell phenotypic transitions: integration of data science and systems biology

Cells with the same genome can exist in different phenotypes and can change between distinct phenotypes when subject to specific stimuli and microenvironments. Some examples include cell differentiation during development, reprogramming for induced pluripotent stem cells and transdifferentiation, cancer metastasis and fibrosis progression. The regulation and dynamics of cell phenotypic conversion is a fundamental problem in biology, and has a long history of being studied within the formalism of dynamical systems. A main challenge for mechanism-driven modeling studies is acquiring sufficient amount of quantitative information for constraining model parameters. Advances in quantitative experimental approaches, especially high throughput single-cell techniques, have accelerated the emergence of a new direction for reconstructing the governing dynamical equations of a cellular system from quantitative single-cell data, beyond the dominant statistical approaches. Here I review a selected number of recent studies using live- and fixed-cell data and provide my perspective on future development.

An Innovative Approach to Tissue Processing and Cell Sorting of Fixed Cells for Subsequent Single-Cell RNA Sequencing

Although single-cell RNA sequencing (scRNA-seq) is currently the gold standard for the analysis of cell-specific expression profiles, the options for processing, staining, and preserving fresh cells remain very limited. Immediate and correct tissue processing is a critical determinant of scRNA-seq success. One major limitation is the restricted compatibility of fixation approaches, which must not destabilize or alter antibody labeling or RNA content or interfere with cell integrity. An additional limitation is the availability of expensive, high-demand cell-sorting equipment to exclude debris and dead or unwanted cells before proceeding with sample sequencing. The goal of this study was to develop a method that allows cells to be fixed and stored prior to FACS sorting for scRNA-seq without compromising the quality of the results. Finally, the challenge of preserving as many living cells as possible during tissue processing is another crucial issue addressed in this study. Our study focused on pancreatic ductal adenocarcinoma samples, where the number of live cells is rather limited, as in many other tumor tissues. Harsh tissue dissociation methods and sample preparation for analysis can negatively affect cell viability. Using the murine pancreatic cancer model Pan02, we evaluated the semi-automated mechanical/enzymatic digestion of solid tumors by gentleMACS Dissociator and compared it with mechanical dissociation of the same tissue. Moreover, we investigated a type of cell fixation that is successful in preserving cell RNA integrity yet compatible with FACS and subsequent scRNA-sequencing. Our protocol allows tissue to be dissociated and stained in one day and proceeds to cell sorting and scRNA-seq later, which is a great advantage for processing clinical patient material.

Wound healing, fibroblast heterogeneity, and fibrosis

Fibroblasts are highly dynamic cells that play a central role in tissue repair and fibrosis. However, the mechanisms by which they contribute to both physiologic and pathologic states of extracellular matrix deposition and remodeling are just starting to be understood. In this review article, we discuss the current state of knowledge in fibroblast biology and heterogeneity, with a primary focus on the role of fibroblasts in skin wound repair. We also consider emerging techniques in the field, which enable an increasingly nuanced and contextualized understanding of these complex systems, and evaluate limitations of existing methodologies and knowledge. Collectively, this review spotlights a diverse body of research examining an often-overlooked cell type-the fibroblast-and its critical functions in wound repair and beyond.

Lessons from single-cell RNA sequencing of human islets

Islet dysfunction is central in type 2 diabetes and full-blown type 2 diabetes develops first when the beta cells lose their ability to secrete adequate amounts of insulin in response to raised plasma glucose. Several mechanisms behind beta cell dysfunction have been put forward but many important questions still remain. Furthermore, our understanding of the contribution of each islet cell type in type 2 diabetes pathophysiology has been limited by technical boundaries. Closing this knowledge gap will lead to a leap forward in our understanding of the islet as an organ and potentially lead to improved treatments. The development of single-cell RNA sequencing (scRNAseq) has led to a breakthrough for characterising the transcriptome of each islet cell type and several important observations on the regulation of cell-type-specific gene expression have been made. When it comes to identifying type 2 diabetes disease mechanisms, the outcome is still limited. Several studies have identified differentially expressed genes, although there is very limited consensus between the studies. As with all new techniques, scRNAseq has limitations; in addition to being extremely expensive, genes expressed at low levels may not be detected, noise may not be appropriately filtered and selection biases for certain cell types are at hand. Furthermore, recent advances suggest that commonly used computational tools may be suboptimal for analysis of scRNAseq data in small-scale studies. Fortunately, development of new computational tools holds promise for harnessing the full potential of scRNAseq data. Here we summarise how scRNAseq has contributed to increasing the understanding of various aspects of islet biology as well as type 2 diabetes disease mechanisms. We also focus on challenges that remain and propose steps to promote the utilisation of the full potential of scRNAseq in this area.

Statistics in everyone’s backyard: An impact study via citation network analysis

Statistical methodologies are indispensable in data-driven scientific discoveries. In this paper, we make the first effort to understand the impact of recent statistical innovations on other scientific fields. By collecting comprehensive bibliometric data from the Web of Science database for selected statistical journals, we investigate the citation trends and compositions of citing fields over time, and we find increasing citation diversity. Furthermore, in a new setting, we apply a local clustering technique involving personalized PageRank with graph conductance for size selection to find the most relevant statistical innovation for a given external topic in other fields. Through a number of case studies, we show that the results from our citation data analysis align well with our knowledge and intuition about these external topics. Overall, we have found that the statistical theory and methods recently invented by the statistics community have made increasing impact on other scientific fields.

•

Citation data were collected for core statistics papers from the past two decades

•

A comprehensive evaluation of their impact on other scientific fields was conducted

•

The external impact of statistics has been increasing in volume and diversity

•

The most influential statistics community for a given external topic was found

How much impact has statistics made on other scientific fields in the era of big data? This work represents the first effort toward quantifying the external influence of statistical theory and method research through citation network analysis. We formulate the problem of finding the most relevant statistical research area for any external research topic as a local clustering problem, suggesting new applied and theoretical grounds for alternative community detection techniques. The results of our analysis confirm that statistics plays an active and expanding role in serving other disciplines. The data we have collected are rich in content and structure, lending themselves naturally to future modeling and analysis from different perspectives.

Forest Fire Clustering for single-cell sequencing combines iterative label propagation with parallelized Monte Carlo simulations

In the era of single-cell sequencing, there is a growing need to extract insights from data with clustering methods. Here, we introduce Forest Fire Clustering, an efficient and interpretable method for cell-type discovery from single-cell data. Forest Fire Clustering makes minimal prior assumptions and, different from current approaches, calculates a non-parametric posterior probability that each cell is assigned a cell-type label. These posterior distributions allow for the evaluation of a label confidence for each cell and enable the computation of “label entropies", highlighting transitions along developmental trajectories. Furthermore, we show that Forest Fire Clustering can make robust, inductive inferences in an online-learning context and can readily scale to millions of cells. Finally, we demonstrate that our method outperforms state-of-the-art clustering approaches on diverse benchmarks of simulated and experimental data. Overall, Forest Fire Clustering is a useful tool for rare cell type discovery in large-scale single-cell analysis.

In the era of single-cell sequencing, there is a growing need to extract insights from data with clustering methods. Here, inspired by forest fire dynamics, the authors devise an algorithm that can cluster single-cell data with minimal prior assumptions and can compute a non-parametric posterior probability for each data point.

From COVID to fibrosis: lessons from single-cell analyses of the human lung

The increased resolution of single-cell RNA-sequencing technologies has led to major breakthroughs and improved our understanding of the normal and pathologic conditions of multiple tissues and organs. In the study of parenchymal lung disease, single-cell RNA-sequencing has better delineated known cell populations and identified novel cells and changes in cellular phenotypes and gene expression patterns associated with disease. In this review, we aim to highlight the advances and insights that have been made possible by applying these technologies to two seemingly very different lung diseases: fibrotic interstitial lung diseases, a group of relentlessly progressive lung diseases leading to pulmonary fibrosis, and COVID-19 pneumonia, an acute viral disease with life-threatening complications, including pulmonary fibrosis. We discuss changes in cell populations and gene expression, highlighting potential common features, such as alveolar cell epithelial injury and aberrant repair and monocyte-derived macrophage populations, as well as relevance and implications to mechanisms of disease and future directions.

Predicting Algorithm of Tissue Cell Ratio Based on Deep Learning Using Single-Cell RNA Sequencing

Background: Understanding the proportion of cell types in heterogeneous tissue samples is important in bioinformatics. It is a challenge to infer the proportion of tissues using bulk RNA sequencing data in bioinformatics because most traditional algorithms for predicting tissue cell ratios heavily rely on standardized specific cell-type gene expression profiles, and do not consider tissue heterogeneity. The prediction accuracy of algorithms is limited, and robustness is lacking. This means that new approaches are needed urgently. Methods: In this study, we introduced an algorithm that automatically predicts tissue cell ratios named Autoptcr. The algorithm uses the data simulated by single-cell RNA sequencing (ScRNA-Seq) for model training, using convolutional neural networks (CNNs) to extract intrinsic relationships between genes and predict the cell proportions of tissues. Results: We trained the algorithm using simulated bulk samples and made predictions using real bulk PBMC data. Comparing Autoptcr with existing advanced algorithms, the Pearson correlation coefficient between the actual value of Autoptcr and the predicted value was the highest, reaching 0.903. Tested on a bulk sample, the correlation coefficient of Lin was 41% higher than that of CSx. The algorithm can infer tissue cell proportions directly from tissue gene expression data. Conclusions: The Autoptcr algorithm uses simulated ScRNA-Seq data for training to solve the problem of specific cell-type gene expression profiles. It also has high prediction accuracy and strong noise resistance for the tissue cell ratio. This work is expected to provide new research ideas for the prediction of tissue cell proportions.

Assessing Algorithmic Thinking Skills in Relation to Age in Early Childhood STEM Education

In the modern digital era, intensive efforts are made to inject computational thinking (CT) across science, technology, engineering, and mathematics (STEM) fields, aiming at formulating a well-trained citizenry and workforce capable of confronting intricate problems that would not be solvable unless exercising CT skills. Focusing on contributing to the research area of CT assessment in the first two years of primary school, we investigated the correlation of algorithmic thinking skills, as a fundamental CT competency, with students’ age in early childhood settings. This article reports a relevant research study, which we implemented under the umbrella of quantitative methodology, employing an innovative assessment tool we constructed for serving the needs of our study. The research was conducted within the context of the environmental study course, adding to the efforts of infusing CT into STEM fields. The study results shed light on the correlation between algorithmic thinking skills and age in early childhood, revealing that age is a predictor factor for algorithmic thinking and, therefore, for CT.

Ly49E separates liver ILC1s into embryo-derived and postnatal subsets with different functions

Type 1 innate lymphoid cells (ILC1s) represent the predominant population of liver ILCs and function as important effectors and regulators of immune responses, but the cellular heterogeneity of ILC1s is not fully understood. Here, single-cell RNA sequencing and flow cytometric analysis demonstrated that liver ILC1s could be dissected into Ly49E⁺ and Ly49E⁻ populations with unique transcriptional and phenotypic features. Genetic fate-mapping analysis revealed that liver Ly49E⁺ ILC1s with strong cytotoxicity originated from embryonic non–bone marrow hematopoietic progenitor cells (HPCs), persisted locally during postnatal life, and mediated protective immunity against cytomegalovirus infection in newborn mice. However, Ly49E⁻ ILC1s developed from BM and extramedullary HPCs after birth, gradually replaced Ly49E⁺ ILC1s in the livers with age, and contained the memory subset in recall response to hapten challenge. Thus, our study shows that Ly49E dissects liver ILC1s into two unique subpopulations, with distinct origins and a bias toward neonatal innate or adult immune memory responses.

Statistical and machine learning methods for spatially resolved transcriptomics data analysis

The recent advancement in spatial transcriptomics technology has enabled multiplexed profiling of cellular transcriptomes and spatial locations. As the capacity and efficiency of the experimental technologies continue to improve, there is an emerging need for the development of analytical approaches. Furthermore, with the continuous evolution of sequencing protocols, the underlying assumptions of current analytical methods need to be re-evaluated and adjusted to harness the increasing data complexity. To motivate and aid future model development, we herein review the recent development of statistical and machine learning methods in spatial transcriptomics, summarize useful resources, and highlight the challenges and opportunities ahead.

The transcriptional portraits of the neural crest at the individual cell level

Since the discovery of this cell population by His in 1850, the neural crest has been under intense study for its important role during vertebrate development. Much has been learned about the function and regulation of neural crest cell differentiation, and as a result, the neural crest has become a key model system for stem cell biology in general. The experiments performed in embryology, genetics, and cell biology in the last 150 years in the neural crest field has given rise to several big questions that have been debated intensely for many years: "How does positional information impact developmental potential? Are neural crest cells individually multipotent or a mixed population of committed progenitors? What are the key factors that regulate the acquisition of stem cell identity, and how does a stem cell decide to differentiate towards one cell fate versus another?" Recently, a marriage between single cell multi-omics, statistical modeling, and developmental biology has shed a substantial amount of light on these questions, and has paved a clear path for future researchers in the field.