Clustering single cells: a review of approaches on high-and low-depth single-cell RNA-seq data

Safe focused ultrasound-mediated blood-brain barrier opening is driven primarily by transient reorganization of tight junctions

Focused ultrasound (FUS) with microbubbles opens the blood-brain barrier (BBB) to allow targeted drug delivery into the brain. The mechanisms by which endothelial cells (ECs) respond to either low acoustic pressures known to open the BBB transiently, or high acoustic pressures that cause brain damage, remain incompletely characterized. Here, we use a mouse strain where tight junctions between ECs are labelled with eGFP and apply FUS at low (450 kPa) and high (750 kPa) acoustic pressures, after which mice are sacrificed at 1 or 72 hours. We find that the EC response leading to FUS-mediated BBB opening at low pressures is localized primarily in arterioles and capillaries, and characterized by a transient loss and reorganization of tight junctions. BBB opening still occurs at low safe pressures in mice lacking caveolae, suggesting that it is driven primarily by transient dismantlement and reorganization of tight junctions. In contrast, BBB opening at high pressures is associated with obliteration of EC tight junctions that remain unrepaired even after 72 hours, allowing continuous fibrinogen passage and persistent microglial activation. Single-cell RNA-sequencing of arteriole, capillary and venule ECs from FUS mice reveals that the transcriptomic responses of ECs exposed to high pressure are dominated by genes belonging to the stress response and cell junction disassembly at both 1 and 72 hours, while lower pressures induce primarily genes responsible for intracellular repair responses in ECs. Our findings suggest that at low pressures transient reorganization of tight junctions and repair responses mediate safe BBB opening for therapeutic delivery.

Significance Statement

Focused ultrasound with microbubbles is used as a noninvasive method to safely open the BBB at low acoustic pressures for therapeutic delivery into the CNS, but the mechanisms mediating this process remain unclear. Kugelman et al., demonstrate that FUS-mediated BBB opening at low pressures occurs primarily in arterioles and capillaries due to transient reorganization of tight junctions. BBB opening still occurs at low safe pressures in mice lacking caveolae, suggesting a transcellular route-independent mechanism. At high unsafe pressures, cell junctions are obliterated and remain unrepaired even after 72 hours, allowing fibrinogen passage and persistent microglial activation. Single-cell RNA-sequencing supports cell biological findings that safe, FUS-mediated BBB opening may be driven by transient reorganization and repair of EC tight junctions.

Implementation and validation of single-cell genomics experiments in neuroscience

Single-cell or single-nucleus transcriptomics is a powerful tool for identifying cell types and cell states. However, hypotheses derived from these assays, including gene expression information, require validation, and their functional relevance needs to be established. The choice of validation depends on numerous factors. Here, we present types of orthogonal and functional validation experiment to strengthen preliminary findings obtained using single-cell and single-nucleus transcriptomics as well as the challenges and limitations of these approaches.

VEGF-A-mediated venous endothelial cell proliferation results in neoangiogenesis during neuroinflammation

Newly formed leaky vessels and blood-brain barrier (BBB) damage are present in demyelinating acute and chronic lesions in multiple sclerosis (MS) and experimental autoimmune encephalomyelitis (EAE). However, the endothelial cell subtypes and signaling pathways contributing to these leaky neovessels are unclear. Here, using single-cell transcriptional profiling and in vivo validation studies, we show that venous endothelial cells express neoangiogenesis gene signatures and show increased proliferation resulting in enlarged veins and higher venous coverage in acute and chronic EAE lesions in female adult mice. These changes correlate with the upregulation of vascular endothelial growth factor A (VEGF-A) signaling. We also confirmed increased expression of neoangiogenic markers in acute and chronic human MS lesions. Treatment with a VEGF-A blocking antibody diminishes the neoangiogenic transcriptomic signatures and vascular proliferation in female adult mice with EAE, but it does not restore BBB function or ameliorate EAE pathology. Our data demonstrate that venous endothelial cells contribute to neoangiogenesis in demyelinating neuroinflammatory conditions.

Identifying cell states in single-cell RNA-seq data at statistically maximal resolution

Single-cell RNA sequencing (scRNA-seq) has become a popular experimental method to study variation of gene expression within a population of cells. However, obtaining an accurate picture of the diversity of distinct gene expression states that are present in a given dataset is highly challenging because of the sparsity of the scRNA-seq data and its inhomogeneous measurement noise properties. Although a vast number of different methods is applied in the literature for clustering cells into subsets with 'similar' expression profiles, these methods generally lack rigorously specified objectives, involve multiple complex layers of normalization, filtering, feature selection, dimensionality-reduction, employ ad hoc measures of distance or similarity between cells, often ignore the known measurement noise properties of scRNA-seq measurements, and include a large number of tunable parameters. Consequently, it is virtually impossible to assign concrete biophysical meaning to the clusterings that result from these methods. Here we address the following problem: Given raw unique molecule identifier (UMI) counts of an scRNA-seq dataset, partition the cells into subsets such that the gene expression states of the cells in each subset are statistically indistinguishable, and each subset corresponds to a distinct gene expression state. That is, we aim to partition cells so as to maximally reduce the complexity of the dataset without removing any of its meaningful structure. We show that, given the known measurement noise structure of scRNA-seq data, this problem is mathematically well-defined and derive its unique solution from first principles. We have implemented this solution in a tool called Cellstates which operates directly on the raw data and automatically determines the optimal partition and cluster number, with zero tunable parameters. We show that, on synthetic datasets, Cellstates almost perfectly recovers optimal partitions. On real data, Cellstates robustly identifies subtle substructure within groups of cells that are traditionally annotated as a common cell type. Moreover, we show that the diversity of gene expression states that Cellstates identifies systematically depends on the tissue of origin and not on technical features of the experiments such as the total number of cells and total UMI count per cell. In addition to the Cellstates tool we also provide a small toolbox of software to place the identified cellstates into a hierarchical tree of higher-order clusters, to identify the most important differentially expressed genes at each branch of this hierarchy, and to visualize these results.

Glutamatergic neuronal activity regulates angiogenesis and blood-retinal barrier maturation via Norrin/β-catenin signaling

Interactions among neuronal, glial, and vascular components are crucial for retinal angiogenesis and blood-retinal barrier (BRB) maturation. Although synaptic dysfunction precedes vascular abnormalities in many retinal pathologies, how neuronal activity, specifically glutamatergic activity, regulates retinal angiogenesis and BRB maturation remains unclear. Using in vivo genetic studies in mice, single-cell RNA sequencing (scRNA-seq), and functional validation, we show that deep plexus angiogenesis and paracellular BRB maturation are delayed in Vglut1-/- retinas where neurons fail to release glutamate. By contrast, deep plexus angiogenesis and paracellular BRB maturation are accelerated in Gnat1-/- retinas, where constitutively depolarized rods release excessive glutamate. Norrin expression and endothelial Norrin/β-catenin signaling are downregulated in Vglut1-/- retinas and upregulated in Gnat1-/- retinas. Pharmacological activation of endothelial Norrin/β-catenin signaling in Vglut1-/- retinas rescues defects in deep plexus angiogenesis and paracellular BRB maturation. Our findings demonstrate that glutamatergic neuronal activity regulates retinal angiogenesis and BRB maturation by modulating endothelial Norrin/β-catenin signaling.

GammaGateR: semi-automated marker gating for single-cell multiplexed imaging

MOTIVATION

Multiplexed immunofluorescence (mIF) is an emerging assay for multichannel protein imaging that can decipher cell-level spatial features in tissues. However, existing automated cell phenotyping methods, such as clustering, face challenges in achieving consistency across experiments and often require subjective evaluation. As a result, mIF analyses often revert to marker gating based on manual thresholding of raw imaging data.

RESULTS

To address the need for an evaluable semi-automated algorithm, we developed GammaGateR, an R package for interactive marker gating designed specifically for segmented cell-level data from mIF images. Based on a novel closed-form gamma mixture model, GammaGateR provides estimates of marker-positive cell proportions and soft clustering of marker-positive cells. The model incorporates user-specified constraints that provide a consistent but slide-specific model fit. We compared GammaGateR against the newest unsupervised approach for annotating mIF data, employing two colon datasets and one ovarian cancer dataset for the evaluation. We showed that GammaGateR produces highly similar results to a silver standard established through manual annotation. Furthermore, we demonstrated its effectiveness in identifying biological signals, achieved by mapping known spatial interactions between CD68 and MUC5AC cells in the colon and by accurately predicting survival in ovarian cancer patients using the phenotype probabilities as input for machine learning methods. GammaGateR is a highly efficient tool that can improve the replicability of marker gating results, while reducing the time of manual segmentation.

AVAILABILITY AND IMPLEMENTATION

The R package is available at https://github.com/JiangmeiRubyXiong/GammaGateR.

Characterization of the responses of brain macrophages to focused ultrasound-mediated blood-brain barrier opening

The opening of the blood-brain barrier (BBB) by focused ultrasound (FUS) coupled with intravenously injected microbubbles can be leveraged as a form of immunotherapy for the treatment of neurodegenerative disorders. However, how FUS BBB opening affects brain macrophages is not well understood. Here by using single-cell sequencing to characterize the distinct responses of microglia and central nervous system-associated macrophages (CAMs) to FUS-mediated BBB opening in mice, we show that the treatment remodels the immune landscape via the recruitment of CAMs and the proliferation of microglia and via population size increases in disease-associated microglia. Both microglia and CAMs showed early and late increases in population sizes, yet only the proliferation of microglia increased at both timepoints. The population of disease-associated microglia also increased, accompanied by the upregulation of genes associated with gliogenesis and phagocytosis, with the depletion of brain macrophages significantly decreasing the duration of BBB opening.

Glutamatergic neuronal activity regulates angiogenesis and blood-retinal barrier maturation via Norrin/β-catenin signaling

Interactions among neuronal, glial and vascular components are crucial for retinal angiogenesis and blood-retinal barrier (BRB) maturation. Although synaptic dysfunction precedes vascular abnormalities in many retinal pathologies, how neuronal activity, specifically glutamatergic activity, regulates retinal angiogenesis and BRB maturation remains unclear. Using in vivo genetic studies in mice, single-cell RNA-sequencing and functional validation, we show that deep plexus angiogenesis and paracellular BRB maturation are delayed in Vglut1 -/- retinas where neurons fail to release glutamate. In contrast, deep plexus angiogenesis and paracellular BRB maturation are accelerated in Gnat1 -/- retinas where constitutively depolarized rods release excessive glutamate. Norrin expression and endothelial Norrin/β-catenin signaling are downregulated in Vglut1 -/- retinas, and upregulated in Gnat1 -/- retinas. Pharmacological activation of endothelial Norrin/β-catenin signaling in Vglut1 -/- retinas rescued defects in deep plexus angiogenesis and paracellular BRB maturation. Our findings demonstrate that glutamatergic neuronal activity regulates retinal angiogenesis and BRB maturation by modulating endothelial Norrin/β-catenin signaling.

CAKE: a flexible self-supervised framework for enhancing cell visualization, clustering and rare cell identification

Single cell sequencing technology has provided unprecedented opportunities for comprehensively deciphering cell heterogeneity. Nevertheless, the high dimensionality and intricate nature of cell heterogeneity have presented substantial challenges to computational methods. Numerous novel clustering methods have been proposed to address this issue. However, none of these methods achieve the consistently better performance under different biological scenarios. In this study, we developed CAKE, a novel and scalable self-supervised clustering method, which consists of a contrastive learning model with a mixture neighborhood augmentation for cell representation learning, and a self-Knowledge Distiller model for the refinement of clustering results. These designs provide more condensed and cluster-friendly cell representations and improve the clustering performance in term of accuracy and robustness. Furthermore, in addition to accurately identifying the major type cells, CAKE could also find more biologically meaningful cell subgroups and rare cell types. The comprehensive experiments on real single-cell RNA sequencing datasets demonstrated the superiority of CAKE in visualization and clustering over other comparison methods, and indicated its extensive application in the field of cell heterogeneity analysis. Contact: Ruiqing Zheng. (rqzheng@csu.edu.cn).

GammaGateR: semi-automated marker gating for single-cell multiplexed imaging

Motivation

Multiplexed immunofluorescence (mIF) is an emerging assay for multichannel protein imaging that can decipher cell-level spatial features in tissues. However, existing automated cell phenotyping methods, such as clustering, face challenges in achieving consistency across experiments and often require subjective evaluation. As a result, mIF analyses often revert to marker gating based on manual thresholding of raw imaging data.

Results

To address the need for an evaluable semi-automated algorithm, we developed GammaGateR, an R package for interactive marker gating designed specifically for segmented cell-level data from mIF images. Based on a novel closed-form gamma mixture model, GammaGateR provides estimates of marker-positive cell proportions and soft clustering of marker-positive cells. The model incorporates user-specified constraints that provide a consistent but slide-specific model fit. We compared GammaGateR against the newest unsupervised approach for annotating mIF data, employing two colon datasets and one ovarian cancer dataset for the evaluation. We showed that GammaGateR produces highly similar results to a silver standard established through manual annotation. Furthermore, we demonstrated its effectiveness in identifying biological signals, achieved by mapping known spatial interactions between CD68 and MUC5AC cells in the colon and by accurately predicting survival in ovarian cancer patients using the phenotype probabilities as input for machine learning methods. GammaGateR is a highly efficient tool that can improve the replicability of marker gating results, while reducing the time of manual segmentation.

Availability and Implementation

The R package is available at https://github.com/JiangmeiRubyXiong/GammaGateR.

FEED: a feature selection method based on gene expression decomposition for single cell clustering

Single-cell clustering is a critical step in biological downstream analysis. The clustering performance could be effectively improved by extracting cell-type-specific genes. The state-of-the-art feature selection methods usually calculate the importance of a single gene without considering the information contained in the gene expression distribution. Moreover, these methods ignore the intrinsic expression patterns of genes and heterogeneity within groups of different mean expression levels. In this work, we present a Feature sElection method based on gene Expression Decomposition (FEED) of scRNA-seq data, which selects informative genes to enhance clustering performance. First, the expression levels of genes are decomposed into multiple Gaussian components. Then, a novel gene correlation calculation method is proposed to measure the relationship between genes from the perspective of distribution. Finally, a permutation-based approach is proposed to determine the threshold of gene importance to obtain marker gene subsets. Compared with state-of-the-art feature selection methods, applying FEED on various scRNA-seq datasets including large datasets followed by different common clustering algorithms results in significant improvements in the accuracy of cell-type identification. The source codes for FEED are freely available at https://github.com/genemine/FEED.

Attention-Based Graph Neural Network for Label Propagation in Single-Cell Omics

Single-cell data analysis has been at forefront of development in biology and medicine since sequencing data have been made available. An important challenge in single-cell data analysis is the identification of cell types. Several methods have been proposed for cell-type identification. However, these methods do not capture the higher-order topological relationship between different samples. In this work, we propose an attention-based graph neural network that captures the higher-order topological relationship between different samples and performs transductive learning for predicting cell types. The evaluation of our method on both simulation and publicly available datasets demonstrates the superiority of our method, scAGN, in terms of prediction accuracy. In addition, our method works best for highly sparse datasets in terms of F1 score, precision score, recall score, and Matthew's correlation coefficients as well. Further, our method's runtime complexity is consistently faster compared to other methods.

ClonoCluster: A method for using clonal origin to inform transcriptome clustering

Clustering cells based on their high-dimensional profiles is an important data reduction process by which researchers infer distinct cellular states. The advent of cellular barcoding, however, provides an alternative means by which to group cells: by their clonal origin. We developed ClonoCluster, a computational method that combines both clone and transcriptome information to create hybrid clusters that weight both kinds of data with a tunable parameter. We generated hybrid clusters across six independent datasets and found that ClonoCluster generated qualitatively different clusters in all cases. The markers of these hybrid clusters were different but had equivalent fidelity to transcriptome-only clusters. The genes most strongly associated with the rearrangements in hybrid clusters were ribosomal function and extracellular matrix genes. We also developed the complementary tool Warp Factor that incorporates clone information in popular 2D visualization techniques like UMAP. Integrating ClonoCluster and Warp Factor revealed biologically relevant markers of cell identity.

Transcriptome analysis of the binucleate ciliate Tetrahymena thermophila with asynchronous nuclear cell cycles

Tetrahymena thermophila harbors two functionally and physically distinct nuclei within a shared cytoplasm. During vegetative growth, the "cell cycles" of the diploid micronucleus and polyploid macronucleus are offset. Micronuclear S phase initiates just before cytokinesis and is completed in daughter cells before onset of macronuclear DNA replication. Mitotic micronuclear division occurs mid-cell cycle, while macronuclear amitosis is coupled to cell division. Here we report the first RNA-seq cell cycle analysis of a binucleated ciliated protozoan. RNA was isolated across 1.5 vegetative cell cycles, starting with a macronuclear G1 population synchronized by centrifugal elutriation. Using MetaCycle, 3244 of the 26,000+ predicted genes were shown to be cell cycle regulated. Proteins present in both nuclei exhibit a single mRNA peak that always precedes their macronuclear function. Nucleus-limited genes, including nucleoporins and importins, are expressed before their respective nucleus-specific role. Cyclin D and A/B gene family members exhibit different expression patterns that suggest nucleus-restricted roles. Periodically expressed genes cluster into seven cyclic patterns. Four clusters have known PANTHER gene ontology terms associated with G1/S and G2/M phase. We propose that these clusters encode known and novel factors that coordinate micro- and macronuclear-specific events such as mitosis, amitosis, DNA replication, and cell division.

Statistical Analysis of Multiplex Immunofluorescence and Immunohistochemistry Imaging Data

Advances in multiplexed single-cell immunofluorescence (mIF) and multiplex immunohistochemistry (mIHC) imaging technologies have enabled the analysis of cell-to-cell spatial relationships that promise to revolutionize our understanding of tissue-based diseases and autoimmune disorders. Multiplex images are collected as multichannel TIFF files; then denoised, segmented to identify cells and nuclei, normalized across slides with protein markers to correct for batch effects, and phenotyped; and then tissue composition and spatial context at the cellular level are analyzed. This chapter discusses methods and software infrastructure for image processing and statistical analysis of mIF/mIHC data.

A Framework for Comparison and Assessment of Synthetic RNA-Seq Data

The ever-growing number of methods for the generation of synthetic bulk and single cell RNA-seq data have multiple and diverse applications. They are often aimed at benchmarking bioinformatics algorithms for purposes such as sample classification, differential expression analysis, correlation and network studies and the optimization of data integration and normalization techniques. Here, we propose a general framework to compare synthetically generated RNA-seq data and select a data-generating tool that is suitable for a set of specific study goals. As there are multiple methods for synthetic RNA-seq data generation, researchers can use the proposed framework to make an informed choice of an RNA-seq data simulation algorithm and software that are best suited for their specific scientific questions of interest.

Identification of adult spinal Shox2 neuronal subpopulations based on unbiased computational clustering of electrophysiological properties

Spinal cord neurons integrate sensory and descending information to produce motor output. The expression of transcription factors has been used to dissect out the neuronal components of circuits underlying behaviors. However, most of the canonical populations of interneurons are heterogeneous and require additional criteria to determine functional subpopulations. Neurons expressing the transcription factor Shox2 can be subclassified based on the co-expression of the transcription factor Chx10 and each subpopulation is proposed to have a distinct connectivity and different role in locomotion. Adult Shox2 neurons have recently been shown to be diverse based on their firing properties. Here, in order to subclassify adult mouse Shox2 neurons, we performed multiple analyses of data collected from whole-cell patch clamp recordings of visually-identified Shox2 neurons from lumbar spinal slices. A smaller set of Chx10 neurons was included in the analyses for validation. We performed k-means and hierarchical unbiased clustering approaches, considering electrophysiological variables. Unlike the categorizations by firing type, the clusters displayed electrophysiological properties that could differentiate between clusters of Shox2 neurons. The presence of clusters consisting exclusively of Shox2 neurons in both clustering techniques suggests that it is possible to distinguish Shox2+Chx10- neurons from Shox2+Chx10+ neurons by electrophysiological properties alone. Computational clusters were further validated by immunohistochemistry with accuracy in a small subset of neurons. Thus, unbiased cluster analysis using electrophysiological properties is a tool that can enhance current interneuronal subclassifications and can complement groupings based on transcription factor and molecular expression.

Gene Co-expression Network Analysis of the Comparative Transcriptome Identifies Hub Genes Associated With Resistance to Aspergillus flavus L. in Cultivated Peanut (Arachis hypogaea L.)

Cultivated peanut (Arachis hypogaea L.), a cosmopolitan oil crop, is susceptible to a variety of pathogens, especially Aspergillus flavus L., which not only vastly reduce the quality of peanut products but also seriously threaten food safety for the contamination of aflatoxin. However, the key genes related to resistance to Aspergillus flavus L. in peanuts remain unclear. This study identifies hub genes positively associated with resistance to A. flavus in two genotypes by comparative transcriptome and weighted gene co-expression network analysis (WGCNA) method. Compared with susceptible genotype (Zhonghua 12, S), the rapid response to A. flavus and quick preparation for the translation of resistance-related genes in the resistant genotype (J-11, R) may be the drivers of its high resistance. WGCNA analysis revealed that 18 genes encoding pathogenesis-related proteins (PR10), 1-aminocyclopropane-1-carboxylate oxidase (ACO1), MAPK kinase, serine/threonine kinase (STK), pattern recognition receptors (PRRs), cytochrome P450, SNARE protein SYP121, pectinesterase, phosphatidylinositol transfer protein, and pentatricopeptide repeat (PPR) protein play major and active roles in peanut resistance to A. flavus. Collectively, this study provides new insight into resistance to A. flavus by employing WGCNA, and the identification of hub resistance-responsive genes may contribute to the development of resistant cultivars by molecular-assisted breeding.

Mouse genomic and cellular annotations

Mice have emerged as one of the most popular and valuable model organisms in the research of human biology. This is due to their genetic and physiological similarity to humans, short generation times, availability of genetically homologous inbred strains, and relatively easy laboratory maintenance. Therefore, following the release of the initial human reference genome, the generation of the mouse reference genome was prioritised and represented an important scientific resource for the mouse genetics community. In 2002, the Mouse Genome Sequencing Consortium published an initial draft of the mouse reference genome which contained ~ 96% of the euchromatic genome of female C57BL/6 J mice. Almost two decades on from the publication of the initial draft, sequencing efforts have continued to increase the completeness and accuracy of the C57BL/6 J reference genome alongside advances in genome annotation. Additionally new sequencing technologies have provided a wealth of data that has added to the repertoire of annotations associated with traditional genomic annotations. Including but not limited to advances in regulatory elements, the 3D genome and individual cellular states. In this review we focus on the reference genome C57BL/6 J and summarise the different aspects of genomic and cellular annotations, as well as their relevance to mouse genetic research. We denote a genomic annotation as a functional unit of the genome. Cellular annotations are annotations of cell type or state, defined by the transcriptomic expression profile of a cell. Due to the wide-ranging number and diversity of annotations describing the mouse genome, we focus on gene, repeat and regulatory element annotation as well as two relatively new technologies; 3D genome architecture and single-cell sequencing outlining their utility in genetic research and their current challenges.

Development of allogeneic HSC-engineered iNKT cells for off-the-shelf cancer immunotherapy

Cell-based immunotherapy has become the new-generation cancer medicine, and "off-the-shelf" cell products that can be manufactured at large scale and distributed readily to treat patients are necessary. Invariant natural killer T (iNKT) cells are ideal cell carriers for developing allogeneic cell therapy because they are powerful immune cells targeting cancers without graft-versus-host disease (GvHD) risk. However, healthy donor blood contains extremely low numbers of endogenous iNKT cells. Here, by combining hematopoietic stem cell (HSC) gene engineering and in vitro differentiation, we generate human allogeneic HSC-engineered iNKT (AlloHSC-iNKT) cells at high yield and purity; these cells closely resemble endogenous iNKT cells, effectively target tumor cells using multiple mechanisms, and exhibit high safety and low immunogenicity. These cells can be further engineered with chimeric antigen receptor (CAR) to enhance tumor targeting or/and gene edited to ablate surface human leukocyte antigen (HLA) molecules and further reduce immunogenicity. Collectively, these preclinical studies demonstrate the feasibility and cancer therapy potential of AlloHSC-iNKT cell products and lay a foundation for their translational and clinical development.

Are dropout imputation methods for scRNA-seq effective for scATAC-seq data?

The tremendous progress of single-cell sequencing technology has given researchers the opportunity to study cell development and differentiation processes at single-cell resolution. Assay of Transposase-Accessible Chromatin by deep sequencing (ATAC-seq) was proposed for genome-wide analysis of chromatin accessibility. Due to technical limitations or other reasons, dropout events are almost a common occurrence for extremely sparse single-cell ATAC-seq data, leading to confusion in downstream analysis (such as clustering). Although considerable progress has been made in the estimation of scRNA-seq data, there is currently no specific method for the inference of dropout events in single-cell ATAC-seq data. In this paper, we select several state-of-the-art scRNA-seq imputation methods (including MAGIC, SAVER, scImpute, deepImpute, PRIME, bayNorm and knn-smoothing) in recent years to infer dropout peaks in scATAC-seq data, and perform a systematic evaluation of these methods through several downstream analyses. Specifically, we benchmarked these methods in terms of correlation with meta-cell, clustering, subpopulations distance analysis, imputation performance for corruption datasets, identification of TF motifs and computation time. The experimental results indicated that most of the imputed peaks increased the correlation with the reference meta-cell, while the performance of different methods on different datasets varied greatly in different downstream analyses, thus should be used with caution. In general, MAGIC performed better than the other methods most consistently across all assessments. Our source code is freely available at https://github.com/yueyueliu/scATAC-master.

Contrastive self-supervised clustering of scRNA-seq data

BACKGROUND

Single-cell RNA sequencing (scRNA-seq) has emerged has a main strategy to study transcriptional activity at the cellular level. Clustering analysis is routinely performed on scRNA-seq data to explore, recognize or discover underlying cell identities. The high dimensionality of scRNA-seq data and its significant sparsity accentuated by frequent dropout events, introducing false zero count observations, make the clustering analysis computationally challenging. Even though multiple scRNA-seq clustering techniques have been proposed, there is no consensus on the best performing approach. On a parallel research track, self-supervised contrastive learning recently achieved state-of-the-art results on images clustering and, subsequently, image classification.

RESULTS

We propose contrastive-sc, a new unsupervised learning method for scRNA-seq data that perform cell clustering. The method consists of two consecutive phases: first, an artificial neural network learns an embedding for each cell through a representation training phase. The embedding is then clustered in the second phase with a general clustering algorithm (i.e. KMeans or Leiden community detection). The proposed representation training phase is a new adaptation of the self-supervised contrastive learning framework, initially proposed for image processing, to scRNA-seq data. contrastive-sc has been compared with ten state-of-the-art techniques. A broad experimental study has been conducted on both simulated and real-world datasets, assessing multiple external and internal clustering performance metrics (i.e. ARI, NMI, Silhouette, Calinski scores). Our experimental analysis shows that constastive-sc compares favorably with state-of-the-art methods on both simulated and real-world datasets.

CONCLUSION

On average, our method identifies well-defined clusters in close agreement with ground truth annotations. Our method is computationally efficient, being fast to train and having a limited memory footprint. contrastive-sc maintains good performance when only a fraction of input cells is provided and is robust to changes in hyperparameters or network architecture. The decoupling between the creation of the embedding and the clustering phase allows the flexibility to choose a suitable clustering algorithm (i.e. KMeans when the number of expected clusters is known, Leiden otherwise) or to integrate the embedding with other existing techniques.

Tutorial: guidelines for annotating single-cell transcriptomic maps using automated and manual methods

Single-cell transcriptomics can profile thousands of cells in a single experiment and identify novel cell types, states and dynamics in a wide variety of tissues and organisms. Standard experimental protocols and analysis workflows have been developed to create single-cell transcriptomic maps from tissues. This tutorial focuses on how to interpret these data to identify cell types, states and other biologically relevant patterns with the objective of creating an annotated map of cells. We recommend a three-step workflow including automatic cell annotation (wherever possible), manual cell annotation and verification. Frequently encountered challenges are discussed, as well as strategies to address them. Guiding principles and specific recommendations for software tools and resources that can be used for each step are covered, and an R notebook is included to help run the recommended workflow. Basic familiarity with computer software is assumed, and basic knowledge of programming (e.g., in the R language) is recommended.

Molecular correlates of muscle spindle and Golgi tendon organ afferents

Proprioceptive feedback mainly derives from groups Ia and II muscle spindle (MS) afferents and group Ib Golgi tendon organ (GTO) afferents, but the molecular correlates of these three afferent subtypes remain unknown. We performed single cell RNA sequencing of genetically identified adult proprioceptors and uncovered five molecularly distinct neuronal clusters. Validation of cluster-specific transcripts in dorsal root ganglia and skeletal muscle demonstrates that two of these clusters correspond to group Ia MS afferents and group Ib GTO afferent proprioceptors, respectively, and suggest that the remaining clusters could represent group II MS afferents. Lineage analysis between proprioceptor transcriptomes at different developmental stages provides evidence that proprioceptor subtype identities emerge late in development. Together, our data provide comprehensive molecular signatures for groups Ia and II MS afferents and group Ib GTO afferents, enabling genetic interrogation of the role of individual proprioceptor subtypes in regulating motor output.

Coordinated movement critically depends on sensory feedback from muscle spindles (MSs) and Golgi tendon organs (GTOs) but the afferents supplying this proprioceptive feedback have remained genetically inseparable. Here the authors use single cell transcriptome analysis to reveal the molecular basis of MS (groups Ia and II) and GTO (group Ib) afferent identities in the mouse.

A hitchhiker's guide to single-cell transcriptomics and data analysis pipelines

Single-cell transcriptomics (SCT) is a tour de force in the era of big omics data that has led to the accumulation of massive cellular transcription data at an astounding resolution of single cells. It provides valuable insights into cells previously unachieved by bulk cell analysis and is proving crucial in uncovering cellular heterogeneity, identifying rare cell populations, distinct cell-lineage trajectories, and mechanisms involved in complex cellular processes. SCT data is highly complex and necessitates advanced statistical and computational methods for analysis. This review provides a comprehensive overview of the steps in a typical SCT workflow, starting from experimental protocol to data analysis, deliberating various pipelines used. We discuss recent trends, challenges, machine learning methods for data analysis, and future prospects. We conclude by listing the multitude of scRNA-seq data applications and how it shall revolutionize our understanding of cellular biology and diseases.

Functional module detection through integration of single-cell RNA sequencing data with protein-protein interaction networks

Background

Recent advances in single-cell RNA sequencing have allowed researchers to explore transcriptional function at a cellular level. In particular, single-cell RNA sequencing reveals that there exist clusters of cells with similar gene expression profiles, representing different transcriptional states.

Results

In this study, we present scPPIN, a method for integrating single-cell RNA sequencing data with protein–protein interaction networks that detects active modules in cells of different transcriptional states. We achieve this by clustering RNA-sequencing data, identifying differentially expressed genes, constructing node-weighted protein–protein interaction networks, and finding the maximum-weight connected subgraphs with an exact Steiner-tree approach. As case studies, we investigate two RNA-sequencing data sets from human liver spheroids and human adipose tissue, respectively. With scPPIN we expand the output of differential expressed genes analysis with information from protein interactions. We find that different transcriptional states have different subnetworks of the protein–protein interaction networks significantly enriched which represent biological pathways. In these pathways, scPPIN identifies proteins that are not differentially expressed but have a crucial biological function (e.g., as receptors) and therefore reveals biology beyond a standard differential expressed gene analysis.

Conclusions

The introduced scPPIN method can be used to systematically analyse differentially expressed genes in single-cell RNA sequencing data by integrating it with protein interaction data. The detected modules that characterise each cluster help to identify and hypothesise a biological function associated to those cells. Our analysis suggests the participation of unexpected proteins in these pathways that are undetectable from the single-cell RNA sequencing data alone. The techniques described here are applicable to other organisms and tissues.

Supplementary Information

The online version contains supplementary material available at (doi:10.1186/s12864-020-07144-2).

A hybrid deep clustering approach for robust cell type profiling using single-cell RNA-seq data

Single-cell RNA sequencing (scRNA-seq) is a recent technology that enables fine-grained discovery of cellular subtypes and specific cell states. Analysis of scRNA-seq data routinely involves machine learning methods, such as feature learning, clustering, and classification, to assist in uncovering novel information from scRNA-seq data. However, current methods are not well suited to deal with the substantial amount of noise that is created by the experiments or the variation that occurs due to differences in the cells of the same type. To address this, we developed a new hybrid approach, deep unsupervised single-cell clustering (DUSC), which integrates feature generation based on a deep learning architecture by using a new technique to estimate the number of latent features, with a model-based clustering algorithm, to find a compact and informative representation of the single-cell transcriptomic data generating robust clusters. We also include a technique to estimate an efficient number of latent features in the deep learning model. Our method outperforms both classical and state-of-the-art feature learning and clustering methods, approaching the accuracy of supervised learning. We applied DUSC to a single-cell transcriptomics data set obtained from a triple-negative breast cancer tumor to identify potential cancer subclones accentuated by copy-number variation and investigate the role of clonal heterogeneity. Our method is freely available to the community and will hopefully facilitate our understanding of the cellular atlas of living organisms as well as provide the means to improve patient diagnostics and treatment.

Impact of similarity metrics on single-cell RNA-seq data clustering

Advances in high-throughput sequencing on single-cell gene expressions [single-cell RNA sequencing (scRNA-seq)] have enabled transcriptome profiling on individual cells from complex samples. A common goal in scRNA-seq data analysis is to discover and characterise cell types, typically through clustering methods. The quality of the clustering therefore plays a critical role in biological discovery. While numerous clustering algorithms have been proposed for scRNA-seq data, fundamentally they all rely on a similarity metric for categorising individual cells. Although several studies have compared the performance of various clustering algorithms for scRNA-seq data, currently there is no benchmark of different similarity metrics and their influence on scRNA-seq data clustering. Here, we compared a panel of similarity metrics on clustering a collection of annotated scRNA-seq datasets. Within each dataset, a stratified subsampling procedure was applied and an array of evaluation measures was employed to assess the similarity metrics. This produced a highly reliable and reproducible consensus on their performance assessment. Overall, we found that correlation-based metrics (e.g. Pearson's correlation) outperformed distance-based metrics (e.g. Euclidean distance). To test if the use of correlation-based metrics can benefit the recently published clustering techniques for scRNA-seq data, we modified a state-of-the-art kernel-based clustering algorithm (SIMLR) using Pearson's correlation as a similarity measure and found significant performance improvement over Euclidean distance on scRNA-seq data clustering. These findings demonstrate the importance of similarity metrics in clustering scRNA-seq data and highlight Pearson's correlation as a favourable choice. Further comparison on different scRNA-seq library preparation protocols suggests that they may also affect clustering performance. Finally, the benchmarking framework is available at http://www.maths.usyd.edu.au/u/SMS/bioinformatics/software.html.

Detecting Interactive Gene Groups for Single-Cell RNA-Seq Data Based on Co-Expression Network Analysis and Subgraph Learning

High-throughput sequencing technologies have enabled the generation of single-cell RNA-seq (scRNA-seq) data, which explore both genetic heterogeneity and phenotypic variation between cells. Some methods have been proposed to detect the related genes causing cell-to-cell variability for understanding tumor heterogeneity. However, most existing methods detect the related genes separately, without considering gene interactions. In this paper, we proposed a novel learning framework to detect the interactive gene groups for scRNA-seq data based on co-expression network analysis and subgraph learning. We first utilized spectral clustering to identify the subpopulations of cells. For each cell subpopulation, the differentially expressed genes were then selected to construct a gene co-expression network. Finally, the interactive gene groups were detected by learning the dense subgraphs embedded in the gene co-expression networks. We applied the proposed learning framework on a real cancer scRNA-seq dataset to detect interactive gene groups of different cancer subtypes. Systematic gene ontology enrichment analysis was performed to examine the detected genes groups by summarizing the key biological processes and pathways. Our analysis shows that different subtypes exhibit distinct gene co-expression networks and interactive gene groups with different functional enrichment. The interactive genes are expected to yield important references for understanding tumor heterogeneity.

scPADGRN: A preconditioned ADMM approach for reconstructing dynamic gene regulatory network using single-cell RNA sequencing data

Disease development and cell differentiation both involve dynamic changes; therefore, the reconstruction of dynamic gene regulatory networks (DGRNs) is an important but difficult problem in systems biology. With recent technical advances in single-cell RNA sequencing (scRNA-seq), large volumes of scRNA-seq data are being obtained for various processes. However, most current methods of inferring DGRNs from bulk samples may not be suitable for scRNA-seq data. In this work, we present scPADGRN, a novel DGRN inference method using “time-series” scRNA-seq data. scPADGRN combines the preconditioned alternating direction method of multipliers with cell clustering for DGRN reconstruction. It exhibits advantages in accuracy, robustness and fast convergence. Moreover, a quantitative index called Differentiation Genes’ Interaction Enrichment (DGIE) is presented to quantify the interaction enrichment of genes related to differentiation. From the DGIE scores of relevant subnetworks, we infer that the functions of embryonic stem (ES) cells are most active initially and may gradually fade over time. The communication strength of known contributing genes that facilitate cell differentiation increases from ES cells to terminally differentiated cells. We also identify several genes responsible for the changes in the DGIE scores occurring during cell differentiation based on three real single-cell datasets. Our results demonstrate that single-cell analyses based on network inference coupled with quantitative computations can reveal key transcriptional regulators involved in cell differentiation and disease development.

Single-cell RNA sequencing (scRNA-seq) data are gaining popularity for providing access to cell-level measurements. Currently, time-series scRNA-seq data allow researchers to study dynamic changes during biological processes. This work proposes a novel method, scPADGRN, for application to time-series scRNA-seq data to construct dynamic gene regulatory networks, which are informative for investigating dynamic changes during disease development and cell differentiation. The proposed method shows satisfactory performance on both simulated data and three real datasets concerning cell differentiation. To quantify network dynamics, we present a quantitative index, DGIE, to measure the degree of activity of a certain set of genes in a regulatory network. Quantitative computations based on dynamic networks identify key regulators in cell differentiation and reveal the activity states of the identified regulators. Specifically, Bhlhe40, Msx2, Foxa2 and Dnmt3l might be important regulatory genes involved in differentiation from mouse ES cells to primitive endoderm (PrE) cells. For differentiation from mouse embryonic fibroblast cells to myocytes, Scx, Fos and Tcf12 are suggested to be key regulators. Sox5, Meis2, Hoxb3, Tcf7l1 and Plagl1 critically contribute during differentiation from human ES cells to definitive endoderm cells. These results may guide further theoretical and experimental efforts to understand cell differentiation processes and explore cell heterogeneity.

Identification of cell types from single cell data using stable clustering

Single-cell RNA-seq (scRNASeq) has become a powerful technique for measuring the transcriptome of individual cells. Unlike the bulk measurements that average the gene expressions over the individual cells, gene measurements at individual cells can be used to study several different tissues and organs at different stages. Identifying the cell types present in the sample from the single cell transcriptome data is a common goal in many single-cell experiments. Several methods have been developed to do this. However, correctly identifying the true cell types remains a challenge. We present a framework that addresses this problem. Our hypothesis is that the meaningful characteristics of the data will remain despite small perturbations of data. We validate the performance of the proposed method on eight publicly available scRNA-seq datasets with known cell types as well as five simulation datasets with different degrees of the cluster separability. We compare the proposed method with five other existing methods: RaceID, SNN-Cliq, SINCERA, SEURAT, and SC3. The results show that the proposed method performs better than the existing methods.

Comparison of high-throughput single-cell RNA sequencing data processing pipelines

With the development of single-cell RNA sequencing (scRNA-seq) technology, it has become possible to perform large-scale transcript profiling for tens of thousands of cells in a single experiment. Many analysis pipelines have been developed for data generated from different high-throughput scRNA-seq platforms, bringing a new challenge to users to choose a proper workflow that is efficient, robust and reliable for a specific sequencing platform. Moreover, as the amount of public scRNA-seq data has increased rapidly, integrated analysis of scRNA-seq data from different sources has become increasingly popular. However, it remains unclear whether such integrated analysis would be biassed if the data were processed by different upstream pipelines. In this study, we encapsulated seven existing high-throughput scRNA-seq data processing pipelines with Nextflow, a general integrative workflow management framework, and evaluated their performance in terms of running time, computational resource consumption and data analysis consistency using eight public datasets generated from five different high-throughput scRNA-seq platforms. Our work provides a useful guideline for the selection of scRNA-seq data processing pipelines based on their performance on different real datasets. In addition, these guidelines can serve as a performance evaluation framework for future developments in high-throughput scRNA-seq data processing.

Single-Cell Sequencing of T cell Receptors: A Perspective on the Technological Development and Translational Application

T cells recognize peptides bound to major histocompatibility complex (MHC) class I and class II molecules at the cell surface. This recognition is accomplished by the expression of T cell receptors (TCR) which are required to be diverse and adaptable in order to accommodate the various and vast number of antigens presented on the MHCs. Thus, determining TCR repertoires of effector T cells is necessary to understand the immunological process in responding to cancer progression, infection, and autoimmune development. Furthermore, understanding the TCR repertoires will provide a solid framework to predict and test the antigen which is more critical in autoimmunity. However, it has been a technical challenge to sequence the TCRs and provide a conceptual context in correlation to the vast number of TCR repertoires in the immunological system. The exploding field of single-cell sequencing has changed how the repertoires are being investigated and analyzed. In this review, we focus on the biology of TCRs, TCR signaling and its implication in autoimmunity. We discuss important methods in bulk sequencing of many cells. Lastly, we explore the most pertinent platforms in single-cell sequencing and its application in autoimmunity.

Species-Specific miRNAs in Human Brain Development and Disease

Identification of the unique features of human brain development and function can be critical towards the elucidation of intricate processes such as higher cognitive functions and human-specific pathologies like neuropsychiatric and behavioral disorders. The developing primate and human central nervous system (CNS) are distinguished by expanded progenitor zones and a protracted time course of neurogenesis, leading to the expansion in brain size, prominent gyral anatomy, distinctive synaptic properties, and complex neural circuits. Comparative genomic studies have revealed that adaptations of brain capacities may be partly explained by human-specific genetic changes that impact the function of proteins associated with neocortical expansion, synaptic function, and language development. However, the formation of complex gene networks may be most relevant for brain evolution. Indeed, recent studies identified distinct human-specific gene expression patterns across developmental time occurring in brain regions linked to cognition. Interestingly, such modules show species-specific divergence and are enriched in genes associated with neuronal development and synapse formation whilst also being implicated in neuropsychiatric diseases. microRNAs represent a powerful component of gene-regulatory networks by promoting spatiotemporal post-transcriptional control of gene expression in the human and primate brain. It has also been suggested that the divergence in miRNA expression plays an important role in shaping gene expression divergence among species. Primate-specific and human-specific miRNAs are principally involved in progenitor proliferation and neurogenic processes but also associate with human cognition, and neurological disorders. Human embryonic or induced pluripotent stem cells and brain organoids, permitting experimental access to neural cells and differentiation stages that are otherwise difficult or impossible to reach in humans, are an essential means for studying species-specific brain miRNAs. Single-cell sequencing approaches can further decode refined miRNA-mRNA interactions during developmental transitions. Elucidating species-specific miRNA regulation will shed new light into the mechanisms that control spatiotemporal events during human brain development and disease, an important step towards fostering novel, holistic and effective therapeutic approaches for neural disorders. In this review, we discuss species-specific regulation of miRNA function, its contribution to the evolving features of the human brain and in neurological disease, with respect also to future therapeutic approaches.

SCALE method for single-cell ATAC-seq analysis via latent feature extraction

Single-cell ATAC-seq (scATAC-seq) profiles the chromatin accessibility landscape at single cell level, thus revealing cell-to-cell variability in gene regulation. However, the high dimensionality and sparsity of scATAC-seq data often complicate the analysis. Here, we introduce a method for analyzing scATAC-seq data, called Single-Cell ATAC-seq analysis via Latent feature Extraction (SCALE). SCALE combines a deep generative framework and a probabilistic Gaussian Mixture Model to learn latent features that accurately characterize scATAC-seq data. We validate SCALE on datasets generated on different platforms with different protocols, and having different overall data qualities. SCALE substantially outperforms the other tools in all aspects of scATAC-seq data analysis, including visualization, clustering, and denoising and imputation. Importantly, SCALE also generates interpretable features that directly link to cell populations, and can potentially reveal batch effects in scATAC-seq experiments.

Clustering methods for single-cell RNA-sequencing expression data: performance evaluation with varying sample sizes and cell compositions

A number of specialized clustering methods have been developed so far for the accurate analysis of single-cell RNA-sequencing (scRNA-seq) expression data, and several reports have been published documenting the performance measures of these clustering methods under different conditions. However, to date, there are no available studies regarding the systematic evaluation of the performance measures of the clustering methods taking into consideration the sample size and cell composition of a given scRNA-seq dataset. Herein, a comprehensive performance evaluation study of 11 selected scRNA-seq clustering methods was performed using synthetic datasets with known sample sizes and number of subpopulations, as well as varying levels of transcriptome complexity. The results indicate that the overall performance of the clustering methods under study are highly dependent on the sample size and complexity of the scRNA-seq dataset. In most of the cases, better clustering performances were obtained as the number of cells in a given expression dataset was increased. The findings of this study also highlight the importance of sample size for the successful detection of rare cell subpopulations with an appropriate clustering tool.

Embracing Systems Toxicology at Single-Cell Resolution

As systems biology expands its multi-omic spectrum to increasing resolutions, distinguishing cells based on single-cell profiles becomes feasible. Unlike traditional bulk assays that average cellular responses and blur the distinct identities of responsive cells, single-cell technologies enable sensitive detection of small cellular changes and precise identification of those cells perturbed by toxicants. Among the suite of omic technologies that continue to expand and become affordable, single-cell RNA sequencing (scRNA-seq) is at the cutting edge and leading the way to transform systems toxicology. Single-cell systems toxicology can provide a wealth of information to elucidate cell-specific alterations and response trajectories, detect points-of-departure, map and develop dynamical models of toxicity pathways.

Single-cell transcriptomics as a framework and roadmap for understanding the brain

A framework for interpreting and guiding experimental examination of the brain is essential for neuroscience. Recently, single-cell RNA sequencing and single-molecule fluorescent in situ hybridization have emerged as key technologies to generate such a framework at a single-cell resolution. These technologies provide a powerful complement for understanding gene expression in the brain: RNA sequencing enables genome-wide high-throughput quantification of gene expression, and in situ hybridization yields spatial registration of gene expression at a cellular resolution. Here, I discuss the insight that each of these technologies individually provide, and how they can be paired in principle and practice to resolve the cell-type-specific spatial organization of the brain. I further discuss the potential of cutting-edge spatial transcriptomics technologies that leverage the advantages of both techniques within the same assay, as well as how transcriptomic assays can be linked with higher-order features of brain structure and function. Such current and forthcoming transcriptomic technologies will have immense impact in generating an underlying logic of the nervous system, and will guide experiments and interpretations across molecular, cellular, circuit, and behavioural neuroscience.

Challenges in unsupervised clustering of single-cell RNA-seq data

Single-cell RNA sequencing (scRNA-seq) allows researchers to collect large catalogues detailing the transcriptomes of individual cells. Unsupervised clustering is of central importance for the analysis of these data, as it is used to identify putative cell types. However, there are many challenges involved. We discuss why clustering is a challenging problem from a computational point of view and what aspects of the data make it challenging. We also consider the difficulties related to the biological interpretation and annotation of the identified clusters.

Molecular Heterogeneity in Large-Scale Biological Data: Techniques and Applications

High-throughput sequencing technologies have evolved at a stellar pace for almost a decade and have greatly advanced our understanding of genome biology. In these sampling-based technologies, there is an important detail that is often overlooked in the analysis of the data and the design of the experiments, specifically that the sampled observations often do not give a representative picture of the underlying population. This has long been recognized as a problem in statistical ecology and in the broader statistics literature. In this review, we discuss the connections between these fields, methodological advances that parallel both the needs and opportunities of large-scale data analysis, and specific applications in modern biology. In the process we describe unique aspects of applying these approaches to sequencing technologies, including sequencing error, population and individual heterogeneity, and the design of experiments.

Single-Cell RNA Sequencing-Based Computational Analysis to Describe Disease Heterogeneity

The trillions of cells in the human body can be viewed as elementary but essential biological units that achieve different body states, but the low resolution of previous cell isolation and measurement approaches limits our understanding of the cell-specific molecular profiles. The recent establishment and rapid growth of single-cell sequencing technology has facilitated the identification of molecular profiles of heterogeneous cells, especially on the transcription level of single cells [single-cell RNA sequencing (scRNA-seq)]. As a novel method, the robustness of scRNA-seq under changing conditions will determine its practical potential in major research programs and clinical applications. In this review, we first briefly presented the scRNA-seq-related methods from the point of view of experiments and computation. Then, we compared several state-of-the-art scRNA-seq analysis frameworks mainly by analyzing their performance robustness on independent scRNA-seq datasets for the same complex disease. Finally, we elaborated on our hypothesis on consensus scRNA-seq analysis and summarized the potential indicative and predictive roles of individual cells in understanding disease heterogeneity by single-cell technologies.

Dual RNA-seq reveals large-scale non-conserved genotype × genotype-specific genetic reprograming and molecular crosstalk in the mycorrhizal symbiosis

Arbuscular mycorrhizal fungi (AMF) impact plant growth and are a major driver of plant diversity and productivity. We quantified the contribution of intra-specific genetic variability in cassava (Manihot esculenta) and Rhizophagus irregularis to gene reprogramming in symbioses using dual RNA-sequencing. A large number of cassava genes exhibited altered transcriptional responses to the fungus but transcription of most of these plant genes (72%) responded in a different direction or magnitude depending on the plant genotype. Two AMF isolates displayed large differences in their transcription, but the direction and magnitude of the transcriptional responses for a large number of these genes was also strongly influenced by the genotype of the plant host. This indicates that unlike the highly conserved plant genes necessary for the symbiosis establishment, most of the plant and fungal gene transcriptional responses are not conserved and are greatly influenced by plant and fungal genetic differences, even at the within-species level. The transcriptional variability detected allowed us to identify an extensive gene network showing the interplay in plant-fungal reprogramming in the symbiosis. Key genes illustrated that the two organisms jointly program their cytoskeleton organization during growth of the fungus inside roots. Our study reveals that plant and fungal genetic variation has a strong role in shaping the genetic reprograming in response to symbiosis, indicating considerable genotype × genotype interactions in the mycorrhizal symbiosis. Such variation needs to be considered in order to understand the molecular mechanisms between AMF and their plant hosts in natural communities.

Identification of cell types in a mouse brain single-cell atlas using low sampling coverage

Background

High throughput methods for profiling the transcriptomes of single cells have recently emerged as transformative approaches for large-scale population surveys of cellular diversity in heterogeneous primary tissues. However, the efficient generation of such atlases will depend on sufficient sampling of diverse cell types while remaining cost-effective to enable a comprehensive examination of organs, developmental stages, and individuals.

Results

To examine the relationship between sampled cell numbers and transcriptional heterogeneity in the context of unbiased cell type classification, we explored the population structure of a publicly available 1.3 million cell dataset from E18.5 mouse brain and validated our findings in published data from adult mice. We propose a computational framework for inferring the saturation point of cluster discovery in a single-cell mRNA-seq experiment, centered around cluster preservation in downsampled datasets. In addition, we introduce a “complexity index,” which characterizes the heterogeneity of cells in a given dataset. Using Cajal-Retzius cells as an example of a limited complexity dataset, we explored whether the detected biological distinctions relate to technical clustering. Surprisingly, we found that clustering distinctions carrying biologically interpretable meaning are achieved with far fewer cells than the originally sampled, though technical saturation of rare populations such as Cajal-Retzius cells is not achieved. We additionally validated these findings with a recently published atlas of cell types across mouse organs and again find using subsampling that a much smaller number of cells recapitulates the cluster distinctions of the complete dataset.

Conclusions

Together, these findings suggest that most of the biologically interpretable cell types from the 1.3 million cell database can be recapitulated by analyzing 50,000 randomly selected cells, indicating that instead of profiling few individuals at high “cellular coverage,” cell atlas studies may instead benefit from profiling more individuals, or many time points at lower cellular coverage and then further enriching for populations of interest. This strategy is ideal for scenarios where cost and time are limited, though extremely rare populations of interest (< 1%) may be identifiable only with much higher cell numbers.

Electronic supplementary material

The online version of this article (10.1186/s12915-018-0580-x) contains supplementary material, which is available to authorized users.

A systematic performance evaluation of clustering methods for single-cell RNA-seq data

Subpopulation identification, usually via some form of unsupervised clustering, is a fundamental step in the analysis of many single-cell RNA-seq data sets. This has motivated the development and application of a broad range of clustering methods, based on various underlying algorithms. Here, we provide a systematic and extensible performance evaluation of 14 clustering algorithms implemented in R, including both methods developed explicitly for scRNA-seq data and more general-purpose methods. The methods were evaluated using nine publicly available scRNA-seq data sets as well as three simulations with varying degree of cluster separability. The same feature selection approaches were used for all methods, allowing us to focus on the investigation of the performance of the clustering algorithms themselves. We evaluated the ability of recovering known subpopulations, the stability and the run time and scalability of the methods. Additionally, we investigated whether the performance could be improved by generating consensus partitions from multiple individual clustering methods. We found substantial differences in the performance, run time and stability between the methods, with SC3 and Seurat showing the most favorable results. Additionally, we found that consensus clustering typically did not improve the performance compared to the best of the combined methods, but that several of the top-performing methods already perform some type of consensus clustering. All the code used for the evaluation is available on GitHub ( https://github.com/markrobinsonuzh/scRNAseq_clustering_comparison). In addition, an R package providing access to data and clustering results, thereby facilitating inclusion of new methods and data sets, is available from Bioconductor ( https://bioconductor.org/packages/DuoClustering2018).

A systematic performance evaluation of clustering methods for single-cell RNA-seq data

Subpopulation identification, usually via some form of unsupervised clustering, is a fundamental step in the analysis of many single-cell RNA-seq data sets. This has motivated the development and application of a broad range of clustering methods, based on various underlying algorithms. Here, we provide a systematic and extensible performance evaluation of 14 clustering algorithms implemented in R, including both methods developed explicitly for scRNA-seq data and more general-purpose methods. The methods were evaluated using nine publicly available scRNA-seq data sets as well as three simulations with varying degree of cluster separability. The same feature selection approaches were used for all methods, allowing us to focus on the investigation of the performance of the clustering algorithms themselves.

We evaluated the ability of recovering known subpopulations, the stability and the run time and scalability of the methods. Additionally, we investigated whether the performance could be improved by generating consensus partitions from multiple individual clustering methods. We found substantial differences in the performance, run time and stability between the methods, with SC3 and Seurat showing the most favorable results. Additionally, we found that consensus clustering typically did not improve the performance compared to the best of the combined methods, but that several of the top-performing methods already perform some type of consensus clustering.

All the code used for the evaluation is available on GitHub ( https://github.com/markrobinsonuzh/scRNAseq_clustering_comparison). In addition, an R package providing access to data and clustering results, thereby facilitating inclusion of new methods and data sets, is available from Bioconductor ( https://bioconductor.org/packages/DuoClustering2018).

A systematic performance evaluation of clustering methods for single-cell RNA-seq data

Subpopulation identification, usually via some form of unsupervised clustering, is a fundamental step in the analysis of many single-cell RNA-seq data sets. This has motivated the development and application of a broad range of clustering methods, based on various underlying algorithms. Here, we provide a systematic and extensible performance evaluation of 14 clustering algorithms implemented in R, including both methods developed explicitly for scRNA-seq data and more general-purpose methods. The methods were evaluated using nine publicly available scRNA-seq data sets as well as three simulations with varying degree of cluster separability. The same feature selection approaches were used for all methods, allowing us to focus on the investigation of the performance of the clustering algorithms themselves. We evaluated the ability of recovering known subpopulations, the stability and the run time and scalability of the methods. Additionally, we investigated whether the performance could be improved by generating consensus partitions from multiple individual clustering methods. We found substantial differences in the performance, run time and stability between the methods, with SC3 and Seurat showing the most favorable results. Additionally, we found that consensus clustering typically did not improve the performance compared to the best of the combined methods, but that several of the top-performing methods already perform some type of consensus clustering. All the code used for the evaluation is available on GitHub ( https://github.com/markrobinsonuzh/scRNAseq_clustering_comparison). In addition, an R package providing access to data and clustering results, thereby facilitating inclusion of new methods and data sets, is available from Bioconductor ( https://bioconductor.org/packages/DuoClustering2018).