Evaluation of single-cell classifiers for single-cell RNA sequencing data sets

scaLR: a low-resource deep neural network-based platform for single cell analysis and biomarker discovery

Single-cell ribonucleic acid (RNA) sequencing (scRNA-seq) produces vast amounts of individual cell profiling data. Its analysis presents a significant challenge in accurately annotating cell types and their associated biomarkers. Different pipelines based on deep neural network (DNN) methods have been employed to tackle these issues. These pipelines have arisen as a promising resource and can extract meaningful and concise features from noisy, diverse, and high-dimensional data to enhance annotations and subsequent analysis. Existing tools require high computational resources to execute large sample datasets. We have developed a cutting-edge platform known as scaLR (Single-cell analysis using low resource) that efficiently processes data into feature subsets, samples in batches to reduce the required memory for processing large datasets, and running DNN models in multiple central processing units. scaLR is equipped with data processing, feature extraction, training, evaluation, and downstream analysis. Its novel feature extraction algorithm first trains the model on a feature subset and stores the importance of the features for all the features in that subset. At the end of the training of all subsets, the top-K features are selected based on their importance. The final model is trained on top-K features; its performance evaluation and associated downstream analysis provide significant biomarkers for different cell types and diseases/traits. Our findings indicate that scaLR offers comparable prediction accuracy and requires less model training time and computational resources than existing Python-based pipelines. We present scaLR, a Python-based platform, engineered to utilize minimal computational resources while maintaining comparable execution times and analysis costs to existing frameworks.

Rhpn2 regulates the development and function of vestibular sensory hair cells through the RhoA signaling in zebrafish

Hearing and balance disorders are significant health issues primarily caused by developmental defects or the irreversible loss of sensory hair cells (HCs). Identifying the underlying genes involved in the morphogenesis and development of HCs is crucial. Our current study highlights rhpn2, a member of rho-binding proteins, as essential for vestibular HC development. The rhpn2 gene is highly expressed in the crista and macula HCs. Loss of rhpn2 function in zebrafish reduces the otic vesicle area and vestibular HC number, accompanied by vestibular dysfunction. Shorter stereocilia and compromised mechanotransduction channel function are found in the crista HCs of rhpn2 mutants. Transcriptome RNA sequencing analysis predicts the potential interaction of rhpn2 with rhoab. Furthermore, co-immunoprecipitation confirms that Rhpn2 directly binds to RhoA, validating the interaction of the two proteins. rhpn2 knockout leads to a decreased expression of rock2b, a canonical RhoA signaling pathway gene. Treatment with the RhoA activator or exogenous rock2b mRNA injection mitigates crista HC stereocilia defects in rhpn2 mutants. This study uncovers the role of rhpn2 in vestibular HC development and stereocilia formation via mediating the RhoA signaling pathway, providing a target for the treatment of balance disorders.

Self-Supervised Graph Representation Learning for Single-Cell Classification

Accurately identifying cell types in single-cell RNA sequencing data is critical for understanding cellular differentiation and pathological mechanisms in downstream analysis. As traditional biological approaches are laborious and time-intensive, it is imperative to develop computational biology methods for cell classification. However, it remains a challenge for existing methods to adequately utilize the potential gene expression information within the vast amount of unlabeled cell data, which limits their classification and generalization performance. Therefore, we propose a novel self-supervised graph representation learning framework for single-cell classification, named scSSGC. Specifically, in the pre-training stage of self-supervised learning, multiple K-means clustering tasks conducted on unlabeled cell data are jointly employed for model training, thereby mitigating the issue of limited labeled data. To effectively capture the potential interactions among cells, we introduce a locally augmented graph neural network to enhance the information aggregation capability for nodes with fewer neighbors in the cell graph. A range of benchmark experiments demonstrates that scSSGC outperforms existing state-of-the-art cell classification methods. More importantly, scSSGC provides stable performance when faced with cross-datasets, indicating better generalization ability.

adverSCarial: assessing the vulnerability of single-cell RNA-sequencing classifiers to adversarial attacks

MOTIVATION

Several machine learning (ML) algorithms dedicated to the detection of healthy and diseased cell types from single-cell RNA sequencing (scRNA-seq) data have been proposed for biomedical purposes. This raises concerns about their vulnerability to adversarial attacks, exploiting threats causing malicious alterations of the classifiers' output with defective and well-crafted input.

RESULTS

With adverSCarial, adversarial attacks of single-cell transcriptomic data can easily be simulated in a range of ways, from expanded but undetectable modifications to aggressive and targeted ones, enabling vulnerability assessment of scRNA-seq classifiers to variations of gene expression, whether technical, biological, or intentional. We exemplify the usefulness and performance with a panel of attack modes proposed in adverSCarial by assessing the robustness of five scRNA-seq classifiers, each belonging to a distinct class of ML algorithm, and explore the potential unlocked by exposing their inner workings and sensitivities on four different datasets. These analyses can guide the development of more reliable models, with improved interpretability, usable in biomedical research and future clinical applications.

AVAILABILITY AND IMPLEMENTATION

adverSCarial is a freely available R package accessible from Bioconductor: https://bioconductor.org/packages/adverSCarial/ or https://doi.org/10.18129/B9.bioc.adverSCarial. A development version is available at https://github.com/GhislainFievet/adverSCarial.

High resolution tissue and cell type identification via single cell transcriptomic profiling

Tissue identification can be instrumental in reconstructing a crime scene but remains a challenging task in forensic investigations. Conventionally, identifying the presence of certain tissue from tissue mixture by predefined cell type markers in bulk fashion is challenging due to limitations in sensitivity and accuracy. In contrast, single-cell RNA sequencing (scRNA-Seq) is a promising technology that has the potential to enhance or even revolutionize tissue and cell type identification. In this study, we developed a high sensitive general purpose single cell annotation pipeline, scTissueID, to accurately evaluate the single cell profile quality and precisely determine the cell and tissue types based on scRNA profiles. By incorporating a crucial and unique reference cell quality differentiation phase of targeting only high confident cells as reference, scTissueID achieved better and consistent performance in determining cell and tissue types compared to 8 state-of-art single cell annotation pipelines and 6 widely adopted machine learning algorithms, as demonstrated through a large-scale and comprehensive comparison study using both forensic-relevant and Human Cell Atlas (HCA) data. We highlighted the significance of cell quality differentiation, a previously undervalued factor. Thus, this study offers a tool capable of accurately and efficiently identifying cell and tissue types, with broad applicability to forensic investigations and other biomedical research endeavors.

scMUG: deep clustering analysis of single-cell RNA-seq data on multiple gene functional modules

Single-cell RNA sequencing (scRNA-seq) has revolutionized our understanding of cellular heterogeneity by providing gene expression data at the single-cell level. Unlike bulk RNA-seq, scRNA-seq allows identification of different cell types within a given tissue, leading to a more nuanced comprehension of cell functions. However, the analysis of scRNA-seq data presents challenges due to its sparsity and high dimensionality. Since bioinformatics plays an important role in the analysis of big data and its utility for the welfare of living beings, it has been widely applied in analyzing scRNA-seq data. To address these challenges, we introduce the scMUG computational pipeline, which incorporates gene functional module information to enhance scRNA-seq clustering analysis. The pipeline includes data preprocessing, cell representation generation, cell-cell similarity matrix construction, and clustering analysis. The scMUG pipeline also introduces a novel similarity measure that combines local density and global distribution in the latent cell representation space. As far as we can tell, this is the first attempt to integrate gene functional associations into scRNA-seq clustering analysis. We curated nine human scRNA-seq datasets to evaluate our scMUG pipeline. With the help of gene functional information and the novel similarity measure, the clustering results from scMUG pipeline present deep insights into functional relationships between gene expression patterns and cellular heterogeneity. In addition, our scMUG pipeline also presents comparable or better clustering performances than other state-of-the-art methods. All source codes of scMUG have been deposited in a GitHub repository with instructions for reproducing all results (https://github.com/degiminnal/scMUG).

SCInter: A comprehensive single-cell transcriptome integration database for human and mouse

Single-cell RNA sequencing (scRNA-seq), which profiles gene expression at the cellular level, has effectively explored cell heterogeneity and reconstructed developmental trajectories. With the increasing research on diseases and biological processes, scRNA-seq datasets are accumulating rapidly, highlighting the urgent need for collecting and processing these data to support comprehensive and effective annotation and analysis. Here, we have developed a comprehensive Single-Cell transcriptome integration database for human and mouse (SCInter, https://bio.liclab.net/SCInter/index.php), which aims to provide a manually curated database that supports the provision of gene expression profiles across various cell types at the sample level. The current version of SCInter includes 115 integrated datasets and 1016 samples, covering nearly 150 tissues/cell lines. It contains 8016,646 cell markers in 457 identified cell types. SCInter enabled comprehensive analysis of cataloged single-cell data encompassing quality control (QC), clustering, cell markers, multi-method cell type automatic annotation, predicting cell differentiation trajectories and so on. At the same time, SCInter provided a user-friendly interface to query, browse, analyze and visualize each integrated dataset and single cell sample, along with comprehensive QC reports and processing results. It will facilitate the identification of cell type in different cell subpopulations and explore developmental trajectories, enhancing the study of cell heterogeneity in the fields of immunology and oncology.

The Role of Immunocyte Infiltration Regulatory Network Based on hdWGCNA and Single-Cell Bioinformatics Analysis in Intervertebral Disc Degeneration

Immune infiltration plays a crucial role in intervertebral disc degeneration (IDD). In this study, we explored the immune microenvironment of IDD through single-cell bioinformatics analysis. Three single-cell datasets were integrated into this study. Nucleus pulposus cells (NPCs) were divided into subgroups based on characteristic genes, and the role of each subgroup in the IDD process was analyzed through pseudo-time trajectory analysis. The hub genes were obtained using hdWGCNA, further identified by bulk datasets and pseudo-time sequence. The expression of the hub genes defined the NPCs related to immune infiltration, and the interaction between these NPCs and immunocytes was explored. The NPCs were divided into four subgroups: reserve NPCs, HCL-NPCs, response NPCs, and support NPCs, which, respectively, dominate the four processes of IDD: non, mild, moderate, and severe degeneration. SPP1 and ICAM1 were identified as the nucleus pulposus immune infiltration hub genes. Macrophages and myelocytes played pro-inflammatory roles in the SPP1-ICAM both-up NPC group through the SPP1-CD44 pathway and ICAM1-ITGB2 ligand-receptor pathway, respectively. At the same time, both-up NPCs sought self-help inflammation remission from neutrophils through the ANXA1-FPR1 pathway. The systematic analysis of the differentiation and immune infiltration landscapes helps to understand IDD's overall development process. Our data suggest that SPP1 and ICAM1 may be new targets for the treatment of inflammatory infiltration in IDD.

CDT1 is a Potential Therapeutic Target for the Progression of NAFLD to HCC and the Exacerbation of Cancer

Aims

This study aimed to identify potential therapeutic targets in the progression from non-alcoholic fatty liver disease (NAFLD) to hepatocellular carcinoma (HCC), with a focus on genes that could influence disease development and progression.

Background

Hepatocellular carcinoma, significantly driven by non-alcoholic fatty liver disease, represents a major global health challenge due to late-stage diagnosis and limited treatment options. This study utilized bioinformatics to analyze data from GEO and TCGA, aiming to uncover molecular biomarkers that bridge NAFLD to HCC. Through identifying critical genes and pathways, our research seeks to advance early detection and develop targeted therapies, potentially improving prognosis and personalizing treatment for NAFLD-HCC patients.

Objectives

Identify key genes that differ between NAFLD and HCC; Analyze these genes to understand their roles in disease progression; Validate the functions of these genes in NAFLD to HCC transition.

Methods

Initially, we identified a set of genes differentially expressed in both NAFLD and HCC using second-generation sequencing data from the GEO and TCGA databases. We then employed a Cox proportional hazards model and a Lasso regression model, applying machine learning techniques to the large sample data from TCGA. This approach was used to screen for key disease-related genes, and an external dataset was utilized for model validation. Additionally, pseudo-temporal sequencing analysis of single-cell sequencing data was performed to further examine the variations in these genes in NAFLD and HCC.

Results

The machine learning analysis identified IGSF3, CENPW, CDT1, and CDC6 as key genes. Furthermore, constructing a machine learning model for CDT1 revealed it to be the most critical gene, with model validation yielding an ROC value greater than 0.80. The single-cell sequencing data analysis confirmed significant variations in the four predicted key genes between the NAFLD and HCC groups.

Conclusion

Our study underscores the pivotal role of CDT1 in the progression from NAFLD to HCC. This finding opens new avenues for early diagnosis and targeted therapy of HCC, highlighting CDT1 as a potential therapeutic target.

Dynamics of alternative polyadenylation in single root cells of Arabidopsis thaliana

Introduction

Single-cell RNA-seq (scRNA-seq) technologies have been widely used to reveal the diversity and complexity of cells, and pioneering studies on scRNA-seq in plants began to emerge since 2019. However, existing studies on plants utilized scRNA-seq focused only on the gene expression regulation. As an essential post-transcriptional mechanism for regulating gene expression, alternative polyadenylation (APA) generates diverse mRNA isoforms with distinct 3' ends through the selective use of different polyadenylation sites in a gene. APA plays important roles in regulating multiple developmental processes in plants, such as flowering time and stress response.

Methods

In this study, we developed a pipeline to identify and integrate APA sites from different scRNA-seq data and analyze APA dynamics in single cells. First, high-confidence poly(A) sites in single root cells were identified and quantified. Second, three kinds of APA markers were identified for exploring APA dynamics in single cells, including differentially expressed poly(A) sites based on APA site expression, APA markers based on APA usages, and APA switching genes based on 3' UTR (untranslated region) length change. Moreover, cell type annotations of single root cells were refined by integrating both the APA information and the gene expression profile.

Results

We comprehensively compiled a single-cell APA atlas from five scRNA-seq studies, covering over 150,000 cells spanning four major tissue branches, twelve cell types, and three developmental stages. Moreover, we quantified the dynamic APA usages in single cells and identified APA markers across tissues and cell types. Further, we integrated complementary information of gene expression and APA profiles to annotate cell types and reveal subtle differences between cell types.

Discussion

This study reveals that APA provides an additional layer of information for determining cell identity and provides a landscape of APA dynamics during Arabidopsis root development.

Ketogenic diet reshapes cancer metabolism through lysine β-hydroxybutyrylation

Lysine β-hydroxybutyrylation (Kbhb) is a post-translational modification induced by the ketogenic diet (KD), a diet showing therapeutic effects on multiple human diseases. Little is known how cellular processes are regulated by Kbhb. Here we show that protein Kbhb is strongly affected by the KD through a multi-omics analysis of mouse livers. Using a small training dataset with known functions, we developed a bioinformatics method for the prediction of functionally important lysine modification sites (pFunK), which revealed functionally relevant Kbhb sites on various proteins, including aldolase B (ALDOB) Lys108. KD consumption or β-hydroxybutyrate supplementation in hepatocellular carcinoma cells increases ALDOB Lys108bhb and inhibits the enzymatic activity of ALDOB. A Kbhb-mimicking mutation (p.Lys108Gln) attenuates ALDOB activity and its binding to substrate fructose-1,6-bisphosphate, inhibits mammalian target of rapamycin signalling and glycolysis, and markedly suppresses cancer cell proliferation. Our study reveals a critical role of Kbhb in regulating cancer cell metabolism and provides a generally applicable algorithm for predicting functionally important lysine modification sites.

CellSTAR: a comprehensive resource for single-cell transcriptomic annotation

Large-scale studies of single-cell sequencing and biological experiments have successfully revealed expression patterns that distinguish different cell types in tissues, emphasizing the importance of studying cellular heterogeneity and accurately annotating cell types. Analysis of gene expression profiles in these experiments provides two essential types of data for cell type annotation: annotated references and canonical markers. In this study, the first comprehensive database of single-cell transcriptomic annotation resource (CellSTAR) was thus developed. It is unique in (a) offering the comprehensive expertly annotated reference data for annotating hundreds of cell types for the first time and (b) enabling the collective consideration of reference data and marker genes by incorporating tens of thousands of markers. Given its unique features, CellSTAR is expected to attract broad research interests from the technological innovations in single-cell transcriptomics, the studies of cellular heterogeneity & dynamics, and so on. It is now publicly accessible without any login requirement at: https://idrblab.org/cellstar.

Peroxisome proliferator-activated receptors signature reveal the head and neck squamous cell carcinoma energy metabolism phenotype and clinical outcome

BACKGROUND

Peroxisome proliferator activating receptors (PPARs) are important regulators of nuclear hormone receptor function, and they play a key role in biological processes such as lipid metabolism, inflammation and cell proliferation. However, their role in head and neck squamous cell carcinoma (HNSC) is unclear.

METHODS

We used multiple datasets, including TCGA-HNSC, GSE41613, GSE139324, PRJEB23709 and IMVigor, to perform a comprehensive analysis of PPAR-related genes in HNSC. Single-cell sequencing data were preprocessed using Seurat packets, and intercellular communication was analyzed using CellChat packets. Functional enrichment analysis of PPAR-related genes was performed using ClusterProfile and GSEA. Prognostic models were constructed using LASSO and Cox regression models, and immunohistochemical analyses were performed using human protein mapping (The Human Protein Atlas).

RESULTS

Our single-cell RNA sequencing analysis revealed distinct cell populations in HNSC, with T cells having the most significant transcriptome differences between tumors and normal tissues. The PPAR features were higher in most cell types in tumor tissues compared with normal tissues. We identified 17 PPAR-associated differentially expressed genes between tumors and normal tissues. A prognostic model based on seven PPAR-associated genes was constructed with high accuracy in predicting 1, 2 and 3 year survival in patients with HNSC. In addition, patients with a low risk score had a higher immune score and a higher proportion of T cells, CD8+ T cells and cytotoxic lymphocytes. They also showed higher immune checkpoint gene expression, suggesting that they might benefit from immunotherapy. PPAR-related genes were found to be closely related to energy metabolism.

CONCLUSIONS

Our study provides a comprehensive understanding of the role of PPAR related genes in HNSC. The identified PPAR features and constructed prognostic models may serve as potential biomarkers for HNSC prognosis and treatment response. In addition, our study found that PPAR-related genes can differentiate energy metabolism and distinguish energy metabolic heterogeneity in HNSC, providing new insights into the molecular mechanisms of HNSC progression and therapeutic response.

Development of a Novel CD8+ T Cell-Associated Signature for Prognostic Assessment in Hepatocellular Carcinoma

OBJECTIVE

The aim of this study was to analyze the clinical significance and prognostic value of CD8+ T cell-related regulatory genes in hepatocellular carcinoma (HCC).

METHODS

This was a retrospective study. We combined TCGA-LIHC and single-cell RNA sequencing data for Lasso-Cox regression analysis to screen for CD8+ T cell-associated genes to construct a novel signature. The expression of the signature genes was detected at cellular and tissue levels using qRT-PCR, immunohistochemistry, and tissue microarrays. The CIBERSORT algorithm was then used to assess the immune microenvironmental differences between the different risk groups and a drug sensitivity analysis was performed to screen for potential HCC therapeutic agents.

RESULTS

An 8-gene CD8 + T cell-associated signature (FABP5, GZMH, ANXA2, KLRB1, CD7, IL7R, BATF, and RGS2) was constructed. Survival analysis showed that high-risk patients had a poorer prognosis in all cohorts. Tumor immune microenvironment analysis revealed 22 immune cell types that differed significantly between patients in different risk groups, with patients in the low-risk group having an immune system that was more active in terms of immune function. Patients in the high-risk group were more prone to immune escape and had a poorer response to immunotherapy, and AZD7762 was screened as the most sensitive drug in the high-risk group. Finally, preliminary experiments have shown that BATF has a promoting effect on the proliferation, migration and invasion of HuH-7 cells.

CONCLUSIONS

The CD8+ T-cell-associated signature is expected to be a tool for optimizing individual patient decision-making and monitoring protocols, and to provide new ideas for treatment and prognostic assessment of HCC.

scPML: pathway-based multi-view learning for cell type annotation from single-cell RNA-seq data

Recent developments in single-cell technology have enabled the exploration of cellular heterogeneity at an unprecedented level, providing invaluable insights into various fields, including medicine and disease research. Cell type annotation is an essential step in its omics research. The mainstream approach is to utilize well-annotated single-cell data to supervised learning for cell type annotation of new singlecell data. However, existing methods lack good generalization and robustness in cell annotation tasks, partially due to difficulties in dealing with technical differences between datasets, as well as not considering the heterogeneous associations of genes in regulatory mechanism levels. Here, we propose the scPML model, which utilizes various gene signaling pathway data to partition the genetic features of cells, thus characterizing different interaction maps between cells. Extensive experiments demonstrate that scPML performs better in cell type annotation and detection of unknown cell types from different species, platforms, and tissues.

Development and validation of a circulating tumor cells-related signature focusing on biochemical recurrence and immunotherapy response in prostate cancer

Background

Studies have shown that the circulating tumor cells (CTCs) play a key role for invasion and formation of distant metastases in prostate cancer (PCa). However, few CTCs-related genes (CRGs) have been developed for biochemical recurrence (BCR) prediction and clinical applications of PCa patients.

Materials and methods

Bioinformatics analysis with public PCa datasets were used to investigate the relationship between the differentially expressed CRGs and BCR. Lasso-COX regression analysis was used to constructed and validated a CRGs-based BCR prediction signature for PCa. Single-cell data were used to validate the expression levels of signature genes in different cell types and then explored the cell-cell communication relationships. Finally, the expression levels of signature genes were verified and the CRGs involved in immunotherapy response were further identified.

Results

Thirteen CRGs were differentially expressed and closely associated with BCR in PCa. Then we constructed and validated a BCR prediction signature for PCa patients based on 3 differentially expressed CRGs (EMID1, SPP1 and UBE2C), and the signature was an independent factor to predict BCR for PCa. Single-cell data showed the specific expression patterns of the signature genes, while the SPP1 pathway plays an important role in cell-cell communication. Further analyses suggested UBE2C was highly expressed in BCR group and high expression of UBE2C had a better response for patients who received immunotherapy. Moreover, the expression levels of UBE2C in CTCs were higher than other cells and tissues, indicated that UBE2C may affect the BCR event of PCa patients through CTCs.

Conclusion

Our findings demonstrated that CRGs were significantly associated with BCR and immunotherapy efficacy in PCa and CRGs may influence the BCR event through CTCs.

Annotation of cell types (ACT): a convenient web server for cell type annotation

BACKGROUND

The advancement of single-cell sequencing has progressed our ability to solve biological questions. Cell type annotation is of vital importance to this process, allowing for the analysis and interpretation of enormous single-cell datasets. At present, however, manual cell annotation which is the predominant approach remains limited by both speed and the requirement of expert knowledge.

METHODS

To address these challenges, we constructed a hierarchically organized marker map through manually curating over 26,000 cell marker entries from about 7000 publications. We then developed WISE, a weighted and integrated gene set enrichment method, to integrate the prevalence of canonical markers and ordered differentially expressed genes of specific cell types in the marker map. Benchmarking analysis suggested that our method outperformed state-of-the-art methods.

RESULTS

By integrating the marker map and WISE, we developed a user-friendly and convenient web server, ACT ( http://xteam.xbio.top/ACT/ or http://biocc.hrbmu.edu.cn/ACT/ ), which only takes a simple list of upregulated genes as input and provides interactive hierarchy maps, together with well-designed charts and statistical information, to accelerate the assignment of cell identities and made the results comparable to expert manual annotation. Besides, a pan-tissue marker map was constructed to assist in cell assignments in less-studied tissues. Applying ACT to three case studies showed that all cell clusters were quickly and accurately annotated, and multi-level and more refined cell types were identified.

CONCLUSIONS

We developed a knowledge-based resource and a corresponding method, together with an intuitive graphical web interface, for cell type annotation. We believe that ACT, emerging as a powerful tool for cell type annotation, would be widely used in single-cell research and considerably accelerate the process of cell type identification.

Construction and validation of a novel prognostic model of neutrophil‑related genes signature of lung adenocarcinoma

Lung adenocarcinoma (LUAD) remains an incurable disease with a poor prognosis. This study aimed to explore neutrophil‑related genes (NRGs) and develop a prognostic signature for predicting the prognosis of LUAD. NRGs were obtained by intersecting modular genes identified by weighted gene co-expression network analysis (WGCNA) using bulk RNA-seq data and the marker genes of neutrophils identified from single-cell RNA-sequencing(scRNA-seq) data. Univariate Cox regression, least absolute shrinkage and selection operator (LASSO), and multivariate Cox analyses were run to construct a prognostic signature, follow by delineation of risk groups, and external validation. Analyses of ESTIMAT, immune function, Tumor Immune Dysfunction and Exclusion (TIDE) scores, Immune cell Proportion Score (IPS), and immune checkpoint genes between high- and low-risk groups were performed, and then analyses of drug sensitivity to screen for sensitive anticancer drugs in high-risk groups. A total of 45 candidate NRGs were identified, of which PLTP, EREG, CD68, CD69, PLAUR, and CYP27A1 were considered to be significantly associated with prognosis in LUAD and were used to construct a prognostic signature. Correlation analysis showed significant differences in the immune landscape between high- and low-risk groups. In addition, our prognostic signature was important for predicting drug sensitivity in the high-risk group. Our study screened for NRGs in LUAD and constructed a novel and effective signature, revealing the immune landscape and providing more appropriate guidance protocols in LUAD treatment.

Shared metabolic shifts in endothelial cells in stroke and Alzheimer's disease revealed by integrated analysis

Since metabolic dysregulation is a hallmark of both stroke and Alzheimer's disease (AD), mining shared metabolic patterns in these diseases will help to identify their possible pathogenic mechanisms and potential intervention targets. However, a systematic integration analysis of the metabolic networks of the these diseases is still lacking. In this study, we integrated single-cell RNA sequencing datasets of ischemic stroke (IS), hemorrhagic stroke (HS) and AD models to construct metabolic flux profiles at the single-cell level. We discovered that the three disorders cause shared metabolic shifts in endothelial cells. These altered metabolic modules were mainly enriched in the transporter-related pathways and were predicted to potentially lead to a decrease in metabolites such as pyruvate and fumarate. We further found that Lef1, Elk3 and Fosl1 may be upstream transcriptional regulators causing metabolic shifts and may be possible targets for interventions that halt the course of neurodegeneration.

GOWDL: gene ontology-driven wide and deep learning model for cell typing of scRNA-seq data

Single-cell RNA-sequencing (scRNA-seq) allows for obtaining genomic and transcriptomic profiles of individual cells. That data make it possible to characterize tissues at the cell level. In this context, one of the main analyses exploiting scRNA-seq data is identifying the cell types within tissue to estimate the quantitative composition of cell populations. Due to the massive amount of available scRNA-seq data, automatic classification approaches for cell typing, based on the most recent deep learning technology, are needed. Here, we present the gene ontology-driven wide and deep learning (GOWDL) model for classifying cell types in several tissues. GOWDL implements a hybrid architecture that considers the functional annotations found in Gene Ontology and the marker genes typical of specific cell types. We performed cross-validation and independent external testing, comparing our algorithm with 12 other state-of-the-art predictors. Classification scores demonstrated that GOWDL reached the best results over five different tissues, except for recall, where we got about 92% versus 97% of the best tool. Finally, we presented a case study on classifying immune cell populations in breast cancer using a hierarchical approach based on GOWDL.

Immune cell type signature discovery and random forest classification for analysis of single cell gene expression datasets

Background

Robust immune cell gene expression signatures are central to the analysis of single cell studies. Nearly all known sets of immune cell signatures have been derived by making use of only single gene expression datasets. Utilizing the power of multiple integrated datasets could lead to high-quality immune cell signatures which could be used as superior inputs to machine learning-based cell type classification approaches.

Results

We established a novel workflow for the discovery of immune cell type signatures based primarily on gene-versus-gene expression similarity. It leverages multiple datasets, here seven single cell expression datasets from six different cancer types and resulted in eleven immune cell type-specific gene expression signatures. We used these to train random forest classifiers for immune cell type assignment for single-cell RNA-seq datasets. We obtained similar or better prediction results compared to commonly used methods for cell type assignment in independent benchmarking datasets. Our gene signature set yields higher prediction scores than other published immune cell type gene sets in random forest-based cell type classification. We further demonstrate how our approach helps to avoid bias in downstream statistical analyses by re-analysis of a published IFN stimulation experiment.

Discussion and conclusion

We demonstrated the quality of our immune cell signatures and their strong performance in a random forest-based cell typing approach. We argue that classifying cells based on our comparably slim sets of genes accompanied by a random forest-based approach not only matches or outperforms widely used published approaches. It also facilitates unbiased downstream statistical analyses of differential gene expression between cell types for significantly more genes compared to previous cell classification algorithms.

Application of single-cell RNA sequencing methods to develop B cell targeted treatments for autoimmunity

The COVID-19 pandemic coincided with several transformative advances in single-cell analysis. These new methods along with decades of research and trials with antibody therapeutics and RNA based technologies allowed for highly effective vaccines and treatments to be produced at astonishing speeds. While these tools were initially focused on models of infection, they also show promise in an autoimmune setting. Self-reactive B cells play important roles as antigen-presenting cells and cytokine and autoantibody producers for many autoimmune diseases. Yet, current therapies to target autoreactive B cells deplete all B cells irrespective of their pathogenicity. Development of self-reactive B cell targeting therapies that would spare non-pathogenic B cells are needed to treat disease while allowing effective immune responses to other ailments. Single-cell RNA sequencing (scRNA-seq) approaches will aid in identification of the pathogenic self-reactive B cells operative in autoimmunity and help with development of more favorable precision targeted therapies.

Development and validation to predict visual acuity and keratometry two years after corneal crosslinking with progressive keratoconus by machine learning

Purpose

To explore and validate the utility of machine learning (ML) methods using a limited sample size to predict changes in visual acuity and keratometry 2 years following corneal crosslinking (CXL) for progressive keratoconus.

Methods

The study included all consecutive patients with progressive keratoconus who underwent CXL from July 2014 to December 2020, with a 2 year follow-up period before July 2022 to develop the model. Variables collected included patient demographics, visual acuity, spherical equivalence, and Pentacam parameters. Available case data were divided into training and testing data sets. Three ML models were evaluated based on their performance in predicting case corrected distance visual acuity (CDVA) and maximum keratometry (Kmax) changes compared to actual values, as indicated by average root mean squared error (RMSE) and R-squared (R2) values. Patients followed from July 2022 to December 2022 were included in the validation set.

Results

A total of 277 eyes from 195 patients were included in training and testing sets and 43 eyes from 35 patients were included in the validation set. The baseline CDVA (26.7%) and the ratio of steep keratometry to flat keratometry (K2/K1; 13.8%) were closely associated with case CDVA changes. The baseline ratio of Kmax to mean keratometry (Kmax/Kmean; 20.9%) was closely associated with case Kmax changes. Using these metrics, the best-performing ML model was XGBoost, which produced predicted values closest to the actual values for both CDVA and Kmax changes in testing set (R2 = 0.9993 and 0.9888) and validation set (R2 = 0.8956 and 0.8382).

Conclusion

Application of a ML approach using XGBoost, and incorporation of identifiable parameters, considerably improved variation prediction accuracy of both CDVA and Kmax 2 years after CXL for treatment of progressive keratoconus.

CD168+ macrophages promote hepatocellular carcinoma tumor stemness and progression through TOP2A/β-catenin/YAP1 axis

Liver cancer stem-like cells (LCSCs) are the main cause of heterogeneity and poor prognosis in hepatocellular carcinoma (HCC). In this study, we aimed to explore the origin of LCSCs and the role of the TOP2A/β-catenin/YAP1 axis in tumor stemness and progression. Using single-cell RNA-seq analysis, we identified TOP2A+CENPF+ LCSCs, which were mainly regulated by CD168+ M2-like macrophages. Furthermore, spatial location analysis and fluorescent staining confirmed that LCSCs were enriched at tumor margins, constituting the spatial heterogeneity of HCC. Mechanistically, TOP2A competitively binds to β-catenin, leading to disassociation of β-catenin from YAP1, promoting HCC stemness and overgrowth. Our study provides valuable insights into the spatial transcriptome heterogeneity of the HCC microenvironment and the critical role of TOP2A/β-catenin/YAP1 axis in HCC stemness and progression.

Somatosensory neuron types and their neural networks as revealed via single-cell transcriptomics

Single-cell RNA sequencing (scRNA-seq) has allowed profiling cell types of the dorsal root ganglia (DRG) and their transcriptional states in physiology and chronic pain. However, the evaluation criteria used in previous studies to classify DRG neurons varied, which presents difficulties in determining the various types of DRG neurons. In this review, we aim to integrate findings from previous transcriptomic studies of the DRG. We first briefly introduce the history of DRG-neuron cell-type profiling, and discuss the advantages and disadvantages of different scRNA-seq methods. We then examine the classification of DRG neurons based on single-cell profiling under physiological and pathological conditions. Finally, we propose further studies on the somatosensory system at the molecular, cellular, and neural network levels.

CIForm as a Transformer-based model for cell-type annotation of large-scale single-cell RNA-seq data

Single-cell omics technologies have made it possible to analyze the individual cells within a biological sample, providing a more detailed understanding of biological systems. Accurately determining the cell type of each cell is a crucial goal in single-cell RNA-seq (scRNA-seq) analysis. Apart from overcoming the batch effects arising from various factors, single-cell annotation methods also face the challenge of effectively processing large-scale datasets. With the availability of an increase in the scRNA-seq datasets, integrating multiple datasets and addressing batch effects originating from diverse sources are also challenges in cell-type annotation. In this work, to overcome the challenges, we developed a supervised method called CIForm based on the Transformer for cell-type annotation of large-scale scRNA-seq data. To assess the effectiveness and robustness of CIForm, we have compared it with some leading tools on benchmark datasets. Through the systematic comparisons under various cell-type annotation scenarios, we exhibit that the effectiveness of CIForm is particularly pronounced in cell-type annotation. The source code and data are available at https://github.com/zhanglab-wbgcas/CIForm.

scRAA: the development of a robust and automatic annotation procedure for single-cell RNA sequencing data

A critical task in single-cell RNA sequencing (scRNA-Seq) data analysis is to identify cell types from heterogeneous tissues. While the majority of classification methods demonstrated high performance in scRNA-Seq annotation problems, a robust and accurate solution is desired to generate reliable outcomes for downstream analyses, for instance, marker genes identification, differentially expressed genes, and pathway analysis. It is hard to establish a universally good metric. Thus, a universally good classification method for all kinds of scenarios does not exist. In addition, reference and query data in cell classification are usually from different experimental batches, and failure to consider batch effects may result in misleading conclusions. To overcome this bottleneck, we propose a robust ensemble approach to classify cells and utilize a batch correction method between reference and query data. We simulated four scenarios that comprise simple to complex batch effect and account for varying cell-type proportions. We further tested our approach on both lung and pancreas data. We found improved prediction accuracy and robust performance across simulation scenarios and real data. The incorporation of batch effect correction between reference and query, and the ensemble approach improve cell-type prediction accuracy while maintaining robustness. We demonstrated these through simulated and real scRNA-Seq data.

Single-cell analysis reveals novel clonally expanded monocytes associated with IL1β-IL1R2 pair in acute inflammatory demyelinating polyneuropathy

Guillain-Barré syndrome (GBS) is an autoimmune disorder wherein the composition and gene expression patterns of peripheral blood immune cells change significantly. It is triggered by antigens with similar epitopes to Schwann cells that stimulate a maladaptive immune response against peripheral nerves. However, an atlas for peripheral blood immune cells in patients with GBS has not yet been constructed. This is a monocentric, prospective study. We collected 5 acute inflammatory demyelinating polyneuropathy (AIDP) patients and 3 healthy controls hospitalized in the First Affiliated Hospital of Harbin Medical University from December 2020 to May 2021, 3 AIDP patients were in the peak stage and 2 were in the convalescent stage. We performed single-cell RNA sequencing (scRNA-seq) of peripheral blood mononuclear cells (PBMCs) from these patients. Furthermore, we performed cell clustering, cell annotation, cell-cell communication, differentially expressed genes (DEGs) identification and pseudotime trajectory analysis. Our study identified a novel clonally expanded CD14+ CD163+ monocyte subtype in the peripheral blood of patients with AIDP, and it was enriched in cellular response to IL1 and chemokine signaling pathways. Furthermore, we observed increased IL1β-IL1R2 cell-cell communication between CD14+ and CD16+ monocytes. In short, by analyzing the single-cell landscape of the PBMCs in patients with AIDP we hope to widen our understanding of the composition of peripheral immune cells in patients with GBS and provide a theoretical basis for future studies.

Evaluation of single-cell RNAseq labelling algorithms using cancer datasets

Single-cell RNA sequencing (scRNA-seq) clustering and labelling methods are used to determine precise cellular composition of tissue samples. Automated labelling methods rely on either unsupervised, cluster-based approaches or supervised, cell-based approaches to identify cell types. The high complexity of cancer poses a unique challenge, as tumor microenvironments are often composed of diverse cell subpopulations with unique functional effects that may lead to disease progression, metastasis and treatment resistance. Here, we assess 17 cell-based and 9 cluster-based scRNA-seq labelling algorithms using 8 cancer datasets, providing a comprehensive large-scale assessment of such methods in a cancer-specific context. Using several performance metrics, we show that cell-based methods generally achieved higher performance and were faster compared to cluster-based methods. Cluster-based methods more successfully labelled non-malignant cell types, likely because of a lack of gene signatures for relevant malignant cell subpopulations. Larger cell numbers present in some cell types in training data positively impacted prediction scores for cell-based methods. Finally, we examined which methods performed favorably when trained and tested on separate patient cohorts in scenarios similar to clinical applications, and which were able to accurately label particularly small or under-represented cell populations in the given datasets. We conclude that scPred and SVM show the best overall performances with cancer-specific data and provide further suggestions for algorithm selection. Our analysis pipeline for assessing the performance of cell type labelling algorithms is available in https://github.com/shooshtarilab/scRNAseq-Automated-Cell-Type-Labelling.

scDLC: a deep learning framework to classify large sample single-cell RNA-seq data

Background

Using single-cell RNA sequencing (scRNA-seq) data to diagnose disease is an effective technique in medical research. Several statistical methods have been developed for the classification of RNA sequencing (RNA-seq) data, including, for example, Poisson linear discriminant analysis (PLDA), negative binomial linear discriminant analysis (NBLDA), and zero-inflated Poisson logistic discriminant analysis (ZIPLDA). Nevertheless, few existing methods perform well for large sample scRNA-seq data, in particular when the distribution assumption is also violated.

Results

We propose a deep learning classifier (scDLC) for large sample scRNA-seq data, based on the long short-term memory recurrent neural networks (LSTMs). Our new scDLC does not require a prior knowledge on the data distribution, but instead, it takes into account the dependency of the most outstanding feature genes in the LSTMs model. LSTMs is a special recurrent neural network, which can learn long-term dependencies of a sequence.

Conclusions

Simulation studies show that our new scDLC performs consistently better than the existing methods in a wide range of settings with large sample sizes. Four real scRNA-seq datasets are also analyzed, and they coincide with the simulation results that our new scDLC always performs the best. The code named “scDLC” is publicly available at https://github.com/scDLC-code/code.

Supplementary Information

The online version contains supplementary material available at (10.1186/s12864-022-08715-1).

Potential biomarkers in the fibrosis progression of nonalcoholic steatohepatitis (NASH)

PURPOSE

Fibrosis is the only histological feature reflecting the severity and prognosis of nonalcoholic steatohepatitis (NASH). We aim to explore novel genes associated with fibrosis progression in NASH.

METHODS

Two human RNA-seq datasets were downloaded from the public database. Weighted gene co-expression network analysis (WGCNA) was used to identify their co-expressed modules and further bioinformatics analysis was performed to identify hub genes within the modules. Finally, based on two single-cell RNA-seq datasets from mice and one microarray dataset from human, we further observed the expression of hub genes in different cell clusters and liver tissues.

RESULTS

7 hub genes (SPP1, PROM1, SOX9, EPCAM, THY1, CD34 and MCAM) associated with fibrosis progression were identified. Single-cell RNA-seq analysis revealed that those hub genes were expressed by different cell clusters such as cholangiocytes, natural killer (NK) cells, and hepatic stellate cells (HSCs). We also found that SPP1 and CD34 serve as markers of different HSCs clusters, which are associated with inflammatory response and fibrogenesis, respectively. Further study suggested that SPP1, SOX9, MCAM and THY1 might be related to NASH-associated hepatocellular carcinoma (HCC). Receiver operating characteristic (ROC) analysis showed that the high expression of these genes could well predict the occurrence of HCC. At the same time, there were significant differences in metabolism-related pathway changes between different HCC subtypes, and SOX9 may be involved in these changes.

CONCLUSIONS

The present study identified novel genes associated with NASH fibrosis and explored their effects on fibrosis from a single-cell perspective that might provide new ideas for the early diagnosis, monitoring, evaluation, and prediction of fibrosis progression in NASH.

Breaking the Immune Complexity of the Tumor Microenvironment Using Single-Cell Technologies

Tumors are not a simple aggregate of transformed cells but rather a complicated ecosystem containing various components, including infiltrating immune cells, tumor-related stromal cells, endothelial cells, soluble factors, and extracellular matrix proteins. Profiling the immune contexture of this intricate framework is now mandatory to develop more effective cancer therapies and precise immunotherapeutic approaches by identifying exact targets or predictive biomarkers, respectively. Conventional technologies are limited in reaching this goal because they lack high resolution. Recent developments in single-cell technologies, such as single-cell RNA transcriptomics, mass cytometry, and multiparameter immunofluorescence, have revolutionized the cancer immunology field, capturing the heterogeneity of tumor-infiltrating immune cells and the dynamic complexity of tenets that regulate cell networks in the tumor microenvironment. In this review, we describe some of the current single-cell technologies and computational techniques applied for immune-profiling the cancer landscape and discuss future directions of how integrating multi-omics data can guide a new "precision oncology" advancement.

scMAGIC: accurately annotating single cells using two rounds of reference-based classification

Here, we introduce scMAGIC (Single Cell annotation using MArker Genes Identification and two rounds of reference-based Classification [RBC]), a novel method that uses well-annotated single-cell RNA sequencing (scRNA-seq) data as the reference to assist in the classification of query scRNA-seq data. A key innovation in scMAGIC is the introduction of a second-round RBC in which those query cells whose cell identities are confidently validated in the first round are used as a new reference to again classify query cells, therefore eliminating the batch effects between the reference and the query data. scMAGIC significantly outperforms 13 competing RBC methods with their optimal parameter settings across 86 benchmark tests, especially when the cell types in the query dataset are not completely covered by the reference dataset and when there exist significant batch effects between the reference and the query datasets. Moreover, when no reference dataset is available, scMAGIC can annotate query cells with reasonably high accuracy by using an atlas dataset as the reference.

Identification of Tumor Mutation Burden, Microsatellite Instability, and Somatic Copy Number Alteration Derived Nine Gene Signatures to Predict Clinical Outcomes in STAD

Genomic features, including tumor mutation burden (TMB), microsatellite instability (MSI), and somatic copy number alteration (SCNA), had been demonstrated to be involved with the tumor microenvironment (TME) and outcome of gastric cancer (GC). We obtained profiles of TMB, MSI, and SCNA by processing 405 GC data from The Cancer Genome Atlas (TCGA) and then conducted a comprehensive analysis though “iClusterPlus.” A total of two subgroups were generated, with distinguished prognosis, somatic mutation burden, copy number changes, and immune landscape. We revealed that Cluster1 was marked by a better prognosis, accompanied by higher TMB, MSIsensor score, TMEscore, and lower SCNA burden. Based on these clusters, we screened 196 differentially expressed genes (DEGs), which were subsequently projected into univariate Cox survival analysis. We constructed a 9-gene immune risk score (IRS) model using LASSO-penalized logistic regression. Moreover, the prognostic prediction of IRS was verified by receiver operating characteristic (ROC) curve analysis and nomogram plot. Another independent Gene Expression Omnibus (GEO) contained specimens from 109 GC patients was designed as an external validation. Our works suggested that the 9‐gene‐signature prediction model, which was derived from TMB, MSI, and SCNA, was a promising predictive tool for clinical outcomes in GC patients. This novel methodology may help clinicians uncover the underlying mechanisms and guide future treatment strategies.

Sincast: a computational framework to predict cell identities in single-cell transcriptomes using bulk atlases as references

Characterizing the molecular identity of a cell is an essential step in single-cell RNA sequencing (scRNA-seq) data analysis. Numerous tools exist for predicting cell identity using single-cell reference atlases. However, many challenges remain, including correcting for inherent batch effects between reference and query data andinsufficient phenotype data from the reference. One solution is to project single-cell data onto established bulk reference atlases to leverage their rich phenotype information. Sincast is a computational framework to query scRNA-seq data by projection onto bulk reference atlases. Prior to projection, single-cell data are transformed to be directly comparable to bulk data, either with pseudo-bulk aggregation or graph-based imputation to address sparse single-cell expression profiles. Sincast avoids batch effect correction, and cell identity is predicted along a continuum to highlight new cell states not found in the reference atlas. In several case study scenarios, we show that Sincast projects single cells into the correct biological niches in the expression space of the bulk reference atlas. We demonstrate the effectiveness of our imputation approach that was specifically developed for querying scRNA-seq data based on bulk reference atlases. We show that Sincast is an efficient and powerful tool for single-cell profiling that will facilitate downstream analysis of scRNA-seq data.

A comprehensive comparison of supervised and unsupervised methods for cell type identification in single-cell RNA-seq

The cell type identification is among the most important tasks in single-cell RNA-sequencing (scRNA-seq) analysis. Many in silico methods have been developed and can be roughly categorized as either supervised or unsupervised. In this study, we investigated the performances of 8 supervised and 10 unsupervised cell type identification methods using 14 public scRNA-seq datasets of different tissues, sequencing protocols and species. We investigated the impacts of a number of factors, including total amount of cells, number of cell types, sequencing depth, batch effects, reference bias, cell population imbalance, unknown/novel cell type, and computational efficiency and scalability. Instead of merely comparing individual methods, we focused on factors' impacts on the general category of supervised and unsupervised methods. We found that in most scenarios, the supervised methods outperformed the unsupervised methods, except for the identification of unknown cell types. This is particularly true when the supervised methods use a reference dataset with high informational sufficiency, low complexity and high similarity to the query dataset. However, such outperformance could be undermined by some undesired dataset properties investigated in this study, which lead to uninformative and biased reference datasets. In these scenarios, unsupervised methods could be comparable to supervised methods. Our study not only explained the cell typing methods' behaviors under different experimental settings but also provided a general guideline for the choice of method according to the scientific goal and dataset properties. Finally, our evaluation workflow is implemented as a modularized R pipeline that allows future evaluation of new methods. Availability: All the source codes are available at https://github.com/xsun28/scRNAIdent.

Vec2image: an explainable artificial intelligence model for the feature representation and classification of high-dimensional biological data by vector-to-image conversion

Feature representation and discriminative learning are proven models and technologies in artificial intelligence fields; however, major challenges for machine learning on large biological datasets are learning an effective model with mechanistical explanation on the model determination and prediction. To satisfy such demands, we developed Vec2image, an explainable convolutional neural network framework for characterizing the feature engineering, feature selection and classifier training that is mainly based on the collaboration of principal component coordinate conversion, deep residual neural networks and embedded k-nearest neighbor representation on pseudo images of high-dimensional biological data, where the pseudo images represent feature measurements and feature associations simultaneously. Vec2image has achieved better performance compared with other popular methods and illustrated its efficiency on feature selection in cell marker identification from tissue-specific single-cell datasets. In particular, in a case study on type 2 diabetes (T2D) by multiple human islet scRNA-seq datasets, Vec2image first displayed robust performance on T2D classification model building across different datasets, then a specific Vec2image model was trained to accurately recognize the cell state and efficiently rank feature genes relevant to T2D which uncovered potential T2D cellular pathogenesis; and next the cell activity changes, cell composition imbalances and cell–cell communication dysfunctions were associated to our finding T2D feature genes from both population-shared and individual-specific perspectives. Collectively, Vec2image is a new and efficient explainable artificial intelligence methodology that can be widely applied in human-readable classification and prediction on the basis of pseudo image representation of biological deep sequencing data.

Cell-type modeling in spatial transcriptomics data elucidates spatially variable colocalization and communication between cell-types in mouse brain

Single-cell spatial transcriptomics (sc-ST) holds the promise to elucidate architectural aspects of complex tissues. Such analyses require modeling cell types in sc-ST datasets through their integration with single-cell RNA-seq datasets. However, this integration, is nontrivial since the two technologies differ widely in the number of profiled genes, and the datasets often do not share many marker genes for given cell types. We developed a neural network model, spatial transcriptomics cell-types assignment using neural networks (STANN), to overcome these challenges. Analysis of STANN's predicted cell types in mouse olfactory bulb (MOB) sc-ST data delineated MOB architecture beyond its morphological layer-based conventional description. We find that cell-type proportions remain consistent within individual morphological layers but vary significantly between layers. Notably, even within a layer, cellular colocalization patterns and intercellular communication mechanisms show high spatial variations. These observations imply a refinement of major cell types into subtypes characterized by spatially localized gene regulatory networks and receptor-ligand usage.

Challenges and Opportunities for the Translation of Single-Cell RNA Sequencing Technologies to Dermatology

Skin is a complex and heterogeneous organ at the cellular level. This complexity is beginning to be understood through the application of single-cell genomics and computational tools. A large number of datasets that shed light on how the different human skin cell types interact in homeostasis-and what ceases to work in diverse dermatological diseases-have been generated and are publicly available. However, translation of these novel aspects to the clinic is lacking. This review aims to summarize the state-of-the-art of skin biology using single-cell technologies, with a special focus on skin pathologies and the translation of mechanistic findings to the clinic. The main implications of this review are to summarize the benefits and limitations of single-cell analysis and thus help translate the emerging insights from these novel techniques to the bedside.

scIAE: an integrative autoencoder-based ensemble classification framework for single-cell RNA-seq data

Single-cell RNA sequencing (scRNA-seq) allows quantitative analysis of gene expression at the level of single cells, beneficial to study cell heterogeneity. The recognition of cell types facilitates the construction of cell atlas in complex tissues or organisms, which is the basis of almost all downstream scRNA-seq data analyses. Using disease-related scRNA-seq data to perform the prediction of disease status can facilitate the specific diagnosis and personalized treatment of disease. Since single-cell gene expression data are high-dimensional and sparse with dropouts, we propose scIAE, an integrative autoencoder-based ensemble classification framework, to firstly perform multiple random projections and apply integrative and devisable autoencoders (integrating stacked, denoising and sparse autoencoders) to obtain compressed representations. Then base classifiers are built on the lower-dimensional representations and the predictions from all base models are integrated. The comparison of scIAE and common feature extraction methods shows that scIAE is effective and robust, independent of the choice of dimension, which is beneficial to subsequent cell classification. By testing scIAE on different types of data and comparing it with existing general and single-cell-specific classification methods, it is proven that scIAE has a great classification power in cell type annotation intradataset, across batches, across platforms and across species, and also disease status prediction. The architecture of scIAE is flexible and devisable, and it is available at https://github.com/JGuan-lab/scIAE.

Besca, a single-cell transcriptomics analysis toolkit to accelerate translational research

Single-cell RNA sequencing (scRNA-seq) revolutionized our understanding of disease biology. The promise it presents to also transform translational research requires highly standardized and robust software workflows. Here, we present the toolkit Besca, which streamlines scRNA-seq analyses and their use to deconvolute bulk RNA-seq data according to current best practices. Beyond a standard workflow covering quality control, filtering, and clustering, two complementary Besca modules, utilizing hierarchical cell signatures and supervised machine learning, automate cell annotation and provide harmonized nomenclatures. Subsequently, the gene expression profiles can be employed to estimate cell type proportions in bulk transcriptomics data. Using multiple, diverse scRNA-seq datasets, some stemming from highly heterogeneous tumor tissue, we show how Besca aids acceleration, interoperability, reusability and interpretability of scRNA-seq data analyses, meeting crucial demands in translational research and beyond.

Single-Cell Analysis for Whole-Organism Datasets

Cell atlases are essential companions to the genome as they elucidate how genes are used in a cell type-specific manner or how the usage of genes changes over the lifetime of an organism. This review explores recent advances in whole-organism single-cell atlases, which enable understanding of cell heterogeneity and tissue and cell fate, both in health and disease. Here we provide an overview of recent efforts to build cell atlases across species and discuss the challenges that the field is currently facing. Moreover, we propose the concept of having a knowledgebase that can scale with the number of experiments and computational approaches and a new feedback loop for development and benchmarking of computational methods that includes contributions from the users. These two aspects are key for community efforts in single-cell biology that will help produce a comprehensive annotated map of cell types and states with unparalleled resolution.

Single-cell classification using graph convolutional networks

BACKGROUND

Analyzing single-cell RNA sequencing (scRNAseq) data plays an important role in understanding the intrinsic and extrinsic cellular processes in biological and biomedical research. One significant effort in this area is the identification of cell types. With the availability of a huge amount of single cell sequencing data and discovering more and more cell types, classifying cells into known cell types has become a priority nowadays. Several methods have been introduced to classify cells utilizing gene expression data. However, incorporating biological gene interaction networks has been proved valuable in cell classification procedures.

RESULTS

In this study, we propose a multimodal end-to-end deep learning model, named sigGCN, for cell classification that combines a graph convolutional network (GCN) and a neural network to exploit gene interaction networks. We used standard classification metrics to evaluate the performance of the proposed method on the within-dataset classification and the cross-dataset classification. We compared the performance of the proposed method with those of the existing cell classification tools and traditional machine learning classification methods.

CONCLUSIONS

Results indicate that the proposed method outperforms other commonly used methods in terms of classification accuracy and F1 scores. This study shows that the integration of prior knowledge about gene interactions with gene expressions using GCN methodologies can extract effective features improving the performance of cell classification.

Taming Cell-to-Cell Heterogeneity in Acute Myeloid Leukaemia With Machine Learning

Acute Myeloid Leukaemia (AML) is a phenotypically and genetically heterogenous blood cancer characterised by very poor prognosis, with disease relapse being the primary cause of treatment failure. AML heterogeneity arise from different genetic and non-genetic sources, including its proposed hierarchical structure, with leukemic stem cells (LSCs) and progenitors giving origin to a variety of more mature leukemic subsets. Recent advances in single-cell molecular and phenotypic profiling have highlighted the intra and inter-patient heterogeneous nature of AML, which has so far limited the success of cell-based immunotherapy approaches against single targets. Machine Learning (ML) can be uniquely used to find non-trivial patterns from high-dimensional datasets and identify rare sub-populations. Here we review some recent ML tools that applied to single-cell data could help disentangle cell heterogeneity in AML by identifying distinct core molecular signatures of leukemic cell subsets. We discuss the advantages and limitations of unsupervised and supervised ML approaches to cluster and classify cell populations in AML, for the identification of biomarkers and the design of personalised therapies.

Single-cell analysis of angiotensin-converting enzyme II expression in human kidneys and bladders reveals a potential route of 2019 novel coronavirus infection

BACKGROUND

Since 2019, a novel coronavirus named 2019 novel coronavirus (2019-nCoV) has emerged worldwide. Apart from fever and respiratory complications, acute kidney injury has been observed in a few patients with coronavirus disease 2019. Furthermore, according to recent findings, the virus has been detected in urine. Angiotensin-converting enzyme II (ACE2) has been proposed to serve as the receptor for the entry of 2019-nCoV, which is the same as that for the severe acute respiratory syndrome. This study aimed to investigate the possible cause of kidney damage and the potential route of 2019-nCoV infection in the urinary system.

METHODS

We used both published kidney and bladder cell atlas data and new independent kidney single-cell RNA sequencing data generated in-house to evaluate ACE2 gene expression in all cell types in healthy kidneys and bladders. The Pearson correlation coefficients between ACE2 and all other genes were first generated. Then, genes with r values larger than 0.1 and P values smaller than 0.01 were deemed significant co-expression genes with ACE2.

RESULTS

Our results showed the enriched expression of ACE2 in all subtypes of proximal tubule (PT) cells of the kidney. ACE2 expression was found in 5.12%, 5.80%, and 14.38% of the proximal convoluted tubule cells, PT cells, and proximal straight tubule cells, respectively, in three published kidney cell atlas datasets. In addition, ACE2 expression was also confirmed in 12.05%, 6.80%, and 10.20% of cells of the proximal convoluted tubule, PT, and proximal straight tubule, respectively, in our own two healthy kidney samples. For the analysis of public data from three bladder samples, ACE2 expression was low but detectable in bladder epithelial cells. Only 0.25% and 1.28% of intermediate cells and umbrella cells, respectively, had ACE2 expression.

CONCLUSION

This study has provided bioinformatics evidence of the potential route of 2019-nCoV infection in the urinary system.

Evaluation of machine learning approaches for cell-type identification from single-cell transcriptomics data

Single-cell transcriptomics is rapidly advancing our understanding of the cellular composition of complex tissues and organisms. A major limitation in most analysis pipelines is the reliance on manual annotations to determine cell identities, which is time-consuming, irreproducible and sometimes lack canonical markers for certain cell types. There is a growing realization of the potential of machine learning models as a supervised classification approach that can significantly aid decision-making processes for cell-type identification. In this work, we performed a comprehensive and impartial evaluation of 10 machine learning models that automatically assign cell phenotypes. The performance of classification methods is estimated by using 20 publicly accessible single-cell RNA sequencing datasets with different sizes, technologies, species and levels of complexity. The performance of each model for within dataset (intra-dataset) and across datasets (inter-dataset) experiments based on the classification accuracy and computation time are both evaluated. Besides, the sensitivity to the number of input features, different annotation levels and dataset complexity was also been estimated. Results showed that most classifiers perform well on a variety of datasets with decreased accuracy for complex datasets, while the Linear Support Vector Machine (linear-SVM) and Logistic Regression classifier models have the best overall performance with remarkably fast computation time. Our work provides a guideline for researchers to select and apply suitable machine learning-based classification models in their analysis workflows and sheds some light on the potential direction of future improvement on automated cell phenotype classification tools based on the single-cell sequencing data.

Automated methods for cell type annotation on scRNA-seq data

The advent of single-cell sequencing started a new era of transcriptomic and genomic research, advancing our knowledge of the cellular heterogeneity and dynamics. Cell type annotation is a crucial step in analyzing single-cell RNA sequencing data, yet manual annotation is time-consuming and partially subjective. As an alternative, tools have been developed for automatic cell type identification. Different strategies have emerged to ultimately associate gene expression profiles of single cells with a cell type either by using curated marker gene databases, correlating reference expression data, or transferring labels by supervised classification. In this review, we present an overview of the available tools and the underlying approaches to perform automated cell type annotations on scRNA-seq data.