Sample-multiplexing approaches for single-cell sequencing

Single-cell sequencing is widely used in biological and medical studies. However, its application with multiple samples is hindered by inefficient sample processing, high experimental costs, ambiguous identification of true single cells, and technical batch effects. Here, we introduce sample-multiplexing approaches for single-cell sequencing in transcriptomics, epigenomics, genomics, and multiomics. In single-cell transcriptomics, sample multiplexing uses variants of native or artificial features as sample markers, enabling sample pooling and decoding. Such features include: (1) natural genetic variation, (2) nucleotide-barcode anchoring on cellular or nuclear membranes, (3) nucleotide-barcode internalization to the cytoplasm or nucleus, (4) vector-based barcode expression in cells, and (5) nucleotide-barcode incorporation during library construction. Other single-cell omics methods are based on similar concepts, particularly single-cell combinatorial indexing. These methods overcome current challenges, while enabling super-loading of single cells. Finally, selection guidelines are presented that can accelerate technological application.

BASS: multi-scale and multi-sample analysis enables accurate cell type clustering and spatial domain detection in spatial transcriptomic studies

Spatial transcriptomic studies are reaching single-cell spatial resolution, with data often collected from multiple tissue sections. Here, we present a computational method, BASS, that enables multi-scale and multi-sample analysis for single-cell resolution spatial transcriptomics. BASS performs cell type clustering at the single-cell scale and spatial domain detection at the tissue regional scale, with the two tasks carried out simultaneously within a Bayesian hierarchical modeling framework. We illustrate the benefits of BASS through comprehensive simulations and applications to three datasets. The substantial power gain brought by BASS allows us to reveal accurate transcriptomic and cellular landscape in both cortex and hypothalamus.

Spatiotemporal Dynamics of the Molecular Expression Pattern and Intercellular Interactions in the Glial Scar Response to Spinal Cord Injury

Nerve regeneration in adult mammalian spinal cord is poor because of the lack of intrinsic regeneration of neurons and extrinsic factors - the glial scar is triggered by injury and inhibits or promotes regeneration. Recent technological advances in spatial transcriptomics (ST) provide a unique opportunity to decipher most genes systematically throughout scar formation, which remains poorly understood. Here, we first constructed the tissue-wide gene expression patterns of mouse spinal cords over the course of scar formation using ST after spinal cord injury from 32 samples. Locally, we profiled gene expression gradients from the leading edge to the core of the scar areas to further understand the scar microenvironment, such as neurotransmitter disorders, activation of the pro-inflammatory response, neurotoxic saturated lipids, angiogenesis, obstructed axon extension, and extracellular structure re-organization. In addition, we described 21 cell transcriptional states during scar formation and delineated the origins, functional diversity, and possible trajectories of subpopulations of fibroblasts, glia, and immune cells. Specifically, we found some regulators in special cell types, such as Thbs1 and Col1a2 in macrophages, CD36 and Postn in fibroblasts, Plxnb2 and Nxpe3 in microglia, Clu in astrocytes, and CD74 in oligodendrocytes. Furthermore, salvianolic acid B, a blood-brain barrier permeation and CD36 inhibitor, was administered after surgery and found to remedy fibrosis. Subsequently, we described the extent of the scar boundary and profiled the bidirectional ligand-receptor interactions at the neighboring cluster boundary, contributing to maintain scar architecture during gliosis and fibrosis, and found that GPR37L1_PSAP, and GPR37_PSAP were the most significant gene-pairs among microglia, fibroblasts, and astrocytes. Last, we quantified the fraction of scar-resident cells and proposed four possible phases of scar formation: macrophage infiltration, proliferation and differentiation of scar-resident cells, scar emergence, and scar stationary. Together, these profiles delineated the spatial heterogeneity of the scar, confirmed the previous concepts about scar architecture, provided some new clues for scar formation, and served as a valuable resource for the treatment of central nervous system injury.

Integrating spatial transcriptomics with single-cell transcriptomics reveals a spatiotemporal gene landscape of the human developing kidney

BACKGROUND

Research on spatiotemporal gene landscape can provide insights into the spatial characteristics of human kidney development and facilitate kidney organoid cultivation. Here, we profiled the spatiotemporal gene programs of the human embryonic kidneys at 9 and 18 post-conception weeks (PCW) by integrating the application of microarray-based spatial transcriptomics and single-cell transcriptomics.

RESULTS

We mapped transcriptomic signatures of scRNA-seq cell types upon the 9 and 18 PCW kidney sections based on cell-type deconvolution and multimodal intersection analyses, depicting a spatial landscape of developing cell subpopulations. We established the gene characteristics in the medullary regions and revealed a strong mitochondrial oxidative phosphorylation and glycolysis activity in the deeper medullary region. We also built a regulatory network centered on GDNF-ETV4 for nephrogenic niche development based on the weighted gene co-expression network analysis and highlighted the key roles of Wnt, FGF, and JAG1-Notch2 signaling in maintaining renal branching morphogenesis.

CONCLUSIONS

Our findings obtained by this spatiotemporal gene program are expected to improve the current understanding of kidney development.

Enabling automated and reproducible spatially resolved transcriptomics at scale

Spatial information of tissues is an essential component to reach a holistic overview of gene expression mechanisms. The sequencing-based Spatial transcriptomics approach allows to spatially barcode the whole transcriptome of tissue sections using microarray glass slides. However, manual preparation of high-quality tissue sequencing libraries is time-consuming and subjected to technical variability. Here, we present an automated adaptation of the 10x Genomics Visium library construction on the widely used Agilent Bravo Liquid Handling Platform. Compared to the manual Visium library preparation, our automated approach reduces hands-on time by over 80% and provides higher throughput and robustness. Our automated Visium library preparation protocol provides a new strategy to standardize spatially resolved transcriptomics analysis of tissues at scale.

Hypoxia classifier for transcriptome datasets

Molecular gene signatures are useful tools to characterize the physiological state of cell populations, but most have developed under a narrow range of conditions and cell types and are often restricted to a set of gene identities. Focusing on the transcriptional response to hypoxia, we aimed to generate widely applicable classifiers sourced from the results of a meta-analysis of 69 differential expression datasets which included 425 individual RNA-seq experiments from 33 different human cell types exposed to different degrees of hypoxia (0.1–5%\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\hbox {O}_{2}$$\end{document}O2) for 2–48 h. The resulting decision trees include both gene identities and quantitative boundaries, allowing for easy classification of individual samples without control or normoxic reference. Each tree is composed of 3–5 genes mostly drawn from a small set of just 8 genes (EGLN1, MIR210HG, NDRG1, ANKRD37, TCAF2, PFKFB3, BHLHE40, and MAFF). In spite of their simplicity, these classifiers achieve over 95% accuracy in cross validation and over 80% accuracy when applied to additional challenging datasets. Our results indicate that the classifiers are able to identify hypoxic tumor samples from bulk RNAseq and hypoxic regions within tumor from spatially resolved transcriptomics datasets. Moreover, application of the classifiers to histological sections from normal tissues suggest the presence of a hypoxic gene expression pattern in the kidney cortex not observed in other normoxic organs. Finally, tree classifiers described herein outperform traditional hypoxic gene signatures when compared against a wide range of datasets. This work describes a set of hypoxic gene signatures, structured as simple decision tress, that identify hypoxic samples and regions with high accuracy and can be applied to a broad variety of gene expression datasets and formats.

Spatial and temporal heterogeneity in the lineage progression of fine oligodendrocyte subtypes

Background

Oligodendrocytes are glial cells that support and insulate axons in the central nervous system through the production of myelin. Oligodendrocytes arise throughout embryonic and early postnatal development from oligodendrocyte precursor cells (OPCs), and recent work demonstrated that they are a transcriptional heterogeneous cell population, but the regional and functional implications of this heterogeneity are less clear. Here, we apply in situ sequencing (ISS) to simultaneously probe the expression of 124 marker genes of distinct oligodendrocyte populations, providing comprehensive maps of the corpus callosum, cingulate, motor, and somatosensory cortex in the brain, as well as gray matter (GM) and white matter (WM) regions in the spinal cord, at postnatal (P10), juvenile (P20), and young adult (P60) stages. We systematically compare the abundances of these populations and investigate the neighboring preference of distinct oligodendrocyte populations.

Results

We observed that oligodendrocyte lineage progression is more advanced in the juvenile spinal cord compared to the brain, corroborating with previous studies. We found myelination still ongoing in the adult corpus callosum while it was more advanced in the cortex. Interestingly, we also observed a lateral-to-medial gradient of oligodendrocyte lineage progression in the juvenile cortex, which could be linked to arealization, as well as a deep-to-superficial gradient with mature oligodendrocytes preferentially accumulating in the deeper layers of the cortex. The ISS experiments also exposed differences in abundances and population dynamics over time between GM and WM regions in the brain and spinal cord, indicating regional differences within GM and WM, and we found that neighboring preferences of some oligodendroglia populations are altered from the juvenile to the adult CNS.

Conclusions

Overall, our ISS experiments reveal spatial heterogeneity of oligodendrocyte lineage progression in the brain and spinal cord and uncover differences in the timing of oligodendrocyte differentiation and myelination, which could be relevant to further investigate functional heterogeneity of oligodendroglia, especially in the context of injury or disease.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12915-022-01325-z.

Modeling zero inflation is not necessary for spatial transcriptomics

BACKGROUND

Spatial transcriptomics are a set of new technologies that profile gene expression on tissues with spatial localization information. With technological advances, recent spatial transcriptomics data are often in the form of sparse counts with an excessive amount of zero values.

RESULTS

We perform a comprehensive analysis on 20 spatial transcriptomics datasets collected from 11 distinct technologies to characterize the distributional properties of the expression count data and understand the statistical nature of the zero values. Across datasets, we show that a substantial fraction of genes displays overdispersion and/or zero inflation that cannot be accounted for by a Poisson model, with genes displaying overdispersion substantially overlapped with genes displaying zero inflation. In addition, we find that either the Poisson or the negative binomial model is sufficient for modeling the majority of genes across most spatial transcriptomics technologies. We further show major sources of overdispersion and zero inflation in spatial transcriptomics including gene expression heterogeneity across tissue locations and spatial distribution of cell types. In particular, when we focus on a relatively homogeneous set of tissue locations or control for cell type compositions, the number of detected overdispersed and/or zero-inflated genes is substantially reduced, and a simple Poisson model is often sufficient to fit the gene expression data there.

CONCLUSIONS

Our study provides the first comprehensive evidence that excessive zeros in spatial transcriptomics are not due to zero inflation, supporting the use of count models without a zero inflation component for modeling spatial transcriptomics.

Spatial transcriptomic analysis reveals inflammatory foci defined by senescent cells in the white matter, hippocampi and cortical grey matter in the aged mouse brain

There is strong evidence that aging is associated with an increased presence of senescent cells in the brain. The finding that treatment with senolytic drugs improves cognitive performance of aged laboratory mice suggests that increased cellular senescence is causally linked to age-related cognitive decline. The relationship between senescent cells and their relative locations within the brain is critical to understanding the pathology of age-related cognitive decline and dementia. To assess spatial distribution of cellular senescence in the aged mouse brain, spatially resolved whole transcriptome mRNA expression was analyzed in sections of brains derived from young (3 months old) and aged (28 months old) C57BL/6 mice while capturing histological information in the same tissue section. Using this spatial transcriptomics (ST)-based method, microdomains containing senescent cells were identified on the basis of their senescence-related gene expression profiles (i.e., expression of the senescence marker cyclin-dependent kinase inhibitor p16INK4A encoded by the Cdkn2a gene) and were mapped to different anatomical brain regions. We confirmed that brain aging is associated with increased cellular senescence in the white matter, the hippocampi and the cortical grey matter. Transcriptional analysis of the senescent cell-containing ST spots shows that presence of senescent cells is associated with a gene expression signature suggestive of neuroinflammation. GO enrichment analysis of differentially expressed genes in the outer region of senescent cell-containing ST spots ("neighboring ST spots") also identified functions related to microglia activation and neuroinflammation. In conclusion, senescent cells accumulate with age in the white matter, the hippocampi and cortical grey matter and likely contribute to the genesis of inflammatory foci in a paracrine manner.

Single-cell and spatially resolved analysis uncovers cell heterogeneity of breast cancer

The heterogeneity and the complex cellular architecture have a crucial effect on breast cancer progression and response to treatment. However, deciphering the neoplastic subtypes and their spatial organization is still challenging. Here, we combine single-nucleus RNA sequencing (snRNA-seq) with a microarray-based spatial transcriptomics (ST) to identify cell populations and their spatial distribution in breast cancer tissues. Malignant cells are clustered into distinct subpopulations. These cell clusters not only have diverse features, origins and functions, but also emerge to the crosstalk within subtypes. Furthermore, we find that these subclusters are mapped in distinct tissue regions, where discrepant enrichment of stromal cell types are observed. We also inferred the abundance of these tumorous subpopulations by deconvolution of large breast cancer RNA-seq cohorts, revealing differential association with patient survival and therapeutic response. Our study provides a novel insight for the cellular architecture of breast cancer and potential therapeutic strategies.

Guidelines for bioinformatics of single-cell sequencing data analysis in Alzheimer's disease: review, recommendation, implementation and application

Alzheimer's disease (AD) is the most common form of dementia, characterized by progressive cognitive impairment and neurodegeneration. Extensive clinical and genomic studies have revealed biomarkers, risk factors, pathways, and targets of AD in the past decade. However, the exact molecular basis of AD development and progression remains elusive. The emerging single-cell sequencing technology can potentially provide cell-level insights into the disease. Here we systematically review the state-of-the-art bioinformatics approaches to analyze single-cell sequencing data and their applications to AD in 14 major directions, including 1) quality control and normalization, 2) dimension reduction and feature extraction, 3) cell clustering analysis, 4) cell type inference and annotation, 5) differential expression, 6) trajectory inference, 7) copy number variation analysis, 8) integration of single-cell multi-omics, 9) epigenomic analysis, 10) gene network inference, 11) prioritization of cell subpopulations, 12) integrative analysis of human and mouse sc-RNA-seq data, 13) spatial transcriptomics, and 14) comparison of single cell AD mouse model studies and single cell human AD studies. We also address challenges in using human postmortem and mouse tissues and outline future developments in single cell sequencing data analysis. Importantly, we have implemented our recommended workflow for each major analytic direction and applied them to a large single nucleus RNA-sequencing (snRNA-seq) dataset in AD. Key analytic results are reported while the scripts and the data are shared with the research community through  GitHub. In summary, this comprehensive review provides insights into various approaches to analyze single cell sequencing data and offers specific guidelines for study design and a variety of analytic directions. The review and the accompanied software tools will serve as a valuable resource for studying cellular and molecular mechanisms of AD, other diseases, or biological systems at the single cell level.

Precision medicine in rheumatoid arthritis

Rheumatoid arthritis is an autoimmune disease that causes significant morbidity. Application of cellular profiling techniques such as single-cell transcriptomics and spatial transcriptomics has uncovered novel pathogenic cell types in RA joint tissues and revealed marked heterogeneity in the cellular composition among RA patients. Together, these insights provide exciting opportunities to translate discoveries into precision medicine in RA. The present review aims to highlight novel insights into RA pathology and discuss key steps needed to translate these discoveries into actionable changes in clinical practice. We review the efforts to identify surrogate biomarkers that could be used to predict RA synovial tissue phenotypes and the corresponding responses to therapy. Finally, we discuss the opportunity to develop novel patient-derived organoid systems as a platform for therapeutic target validation.

Spatial transcriptomic and single-nucleus analysis reveals heterogeneity in a gigantic single-celled syncytium

In multicellular organisms, the specification, coordination, and compartmentalization of cell types enable the formation of complex body plans. However, some eukaryotic protists such as slime molds generate diverse and complex structures while remaining in a multinucleate syncytial state. It is unknown if different regions of these giant syncytial cells have distinct transcriptional responses to environmental encounters and if nuclei within the cell diversify into heterogeneous states. Here, we performed spatial transcriptome analysis of the slime mold Physarum polycephalum in the plasmodium state under different environmental conditions and used single-nucleus RNA-sequencing to dissect gene expression heterogeneity among nuclei. Our data identifies transcriptome regionality in the organism that associates with proliferation, syncytial substructures, and localized environmental conditions. Further, we find that nuclei are heterogenous in their transcriptional profile and may process local signals within the plasmodium to coordinate cell growth, metabolism, and reproduction. To understand how nuclei variation within the syncytium compares to heterogeneity in single-nucleus cells, we analyzed states in single Physarum amoebal cells. We observed amoebal cell states at different stages of mitosis and meiosis, and identified cytokinetic features that are specific to nuclei divisions within the syncytium. Notably, we do not find evidence for predefined transcriptomic states in the amoebae that are observed in the syncytium. Our data shows that a single-celled slime mold can control its gene expression in a region-specific manner while lacking cellular compartmentalization and suggests that nuclei are mobile processors facilitating local specialized functions. More broadly, slime molds offer the extraordinary opportunity to explore how organisms can evolve regulatory mechanisms to divide labor, specialize, balance competition with cooperation, and perform other foundational principles that govern the logic of life.

Advancing root developmental research through single-cell technologies

Single-cell RNA-sequencing has greatly increased the spatiotemporal resolution of root transcriptomics data, but we are still only scratching the surface of its full potential. Despite the challenges that remain in the field, the orderly aligned structure of the Arabidopsis root meristem makes it specifically suitable for lineage tracing and trajectory analysis. These methods will become even more potent by increasing resolution and specificity using tissue-specific single-cell RNA-sequencing and spatial transcriptomics. Feeding multiple single-cell omics data sets into single-cell gene regulatory networks will accelerate the discovery of regulators of root development in multiple species. By providing transcriptome atlases for virtually any species, single-cell technologies could tempt many root developmental biologists to move beyond the comfort of the well-known Arabidopsis root meristem.

Spatially resolved transcriptomics in immersive environments

Spatially resolved transcriptomics is an emerging class of high-throughput technologies that enable biologists to systematically investigate the expression of genes along with spatial information. Upon data acquisition, one major hurdle is the subsequent interpretation and visualization of the datasets acquired. To address this challenge, VR-Cardiomics is presented, which is a novel data visualization system with interactive functionalities designed to help biologists interpret spatially resolved transcriptomic datasets. By implementing the system in two separate immersive environments, fish tank virtual reality (FTVR) and head-mounted display virtual reality (HMD-VR), biologists can interact with the data in novel ways not previously possible, such as visually exploring the gene expression patterns of an organ, and comparing genes based on their 3D expression profiles. Further, a biologist-driven use-case is presented, in which immersive environments facilitate biologists to explore and compare the heart expression profiles of different genes.

Analyzing the Composition and Organization of Ribonucleoprotein Complexes by APEX-Seq

Diverse protein-RNA complexes assemble in cells, and their composition and localization regulate the fate of mRNAs. Here, we detail APEX-Seq, an experimental strategy to capture protein-RNA interactions and profile their sub-cellular organization by in vivo proximity labeling and high-throughput sequencing. APEX-Seq relies on direct proximity labeling of RNAs by the peroxidase enzyme APEX2, which can be targeted to specific sites in the cell or fused to proteins of interest. Direct RNA proximity labeling promises new insights into the dynamic behavior of RNA, addressing length scales beyond direct physical contact but too short for microscopy. APEX-Seq should be widely applicable to diverse biological questions and in many cell types, enabling comprehensive studies of the spatial transcriptome and its dynamics over time.

Epigenetic-Mediated Regulation of Gene Expression for Biological Control and Cancer: Cell and Tissue Structure, Function, and Phenotype

Epigenetic gene regulatory mechanisms play a central role in the biological control of cell and tissue structure, function, and phenotype. Identification of epigenetic dysregulation in cancer provides mechanistic into tumor initiation and progression and may prove valuable for a variety of clinical applications. We present an overview of epigenetically driven mechanisms that are obligatory for physiological regulation and parameters of epigenetic control that are modified in tumor cells. The interrelationship between nuclear structure and function is not mutually exclusive but synergistic. We explore concepts influencing the maintenance of chromatin structures, including phase separation, recognition signals, factors that mediate enhancer-promoter looping, and insulation and how these are altered during the cell cycle and in cancer. Understanding how these processes are altered in cancer provides a potential for advancing capabilities for the diagnosis and identification of novel therapeutic targets.

Insights from Studies of White Adipose Tissue Using Single-Cell Approaches

Technologies allowing studies at single-cell resolution have provided important insights into how different cell populations contribute to tissue function. Application of these methods to white adipose tissue (WAT) has revealed how various metabolic aspects of this organ, such as insulin response, inflammation and tissue expansion, are regulated by specific WAT resident cells, including different subtypes of adipocytes, adipocyte progenitors as well as immune and endothelial cells. In this chapter, we provide an overview of the different technical approaches, their strengths and weaknesses, and summarize how these studies have improved our understanding of WAT function in health and disease.

Optical Cell Tagging for Spatially Resolved Single-Cell RNA Sequencing

Single-cell RNA sequencing (scRNA-seq) has emerged as a powerful tool for profiling gene expression of distinct cell populations at the single-cell level. However, the information of the positions of cells within the multicellular samples is missing in scRNA-seq datasets. To overcome this limitation, we herein develop OpTAG (optical cell tagging) as a new chemical platform for attaching functional tags onto cell surfaces in a spatially resolved manner. With OpTAG, we establish OpTAG-seq, which enables spatially resolved scRNA-seq. We apply OpTAG-seq to investigate the spatially defined transcriptional program in migrating cancer cells and identified a list of genes that are potential regulators for cancer cell migration and invasion. OpTAG-seq provides a convenient method for mapping cellular heterogeneity with spatial information within multicellular biological systems.

Neuropsychiatric disorders: An immunological perspective

Neuropsychiatric diseases have traditionally been studied from brain, and mind-centric perspectives. However, mounting epidemiological and clinical evidence shows a strong correlation of neuropsychiatric manifestations with immune system activation, suggesting a likely mechanistic interaction between the immune and nervous systems in mediating neuropsychiatric disease. Indeed, immune mediators such as cytokines, antibodies, and complement proteins have been shown to affect various cellular members of the central nervous system in multitudinous ways, such as by modulating neuronal firing rates, inducing cellular apoptosis, or triggering synaptic pruning. These observations have in turn led to the exciting development of clinical therapies aiming to harness this neuro-immune interaction for the treatment of neuropsychiatric disease and symptoms. Besides the clinic, important theoretical fundamentals can be drawn from the immune system and applied to our understanding of the brain and neuropsychiatric disease. These new frameworks could lead to novel insights in the field and further potentiate the development of future therapies to treat neuropsychiatric disease.

Novel omics technology driving translational research in precision oncology

In this review, we summarize the current challenges faced by cancer researchers and motivate the use of novel genomics solutions. We follow this up with a comprehensive overview of three recent genomics technologies: liquid biopsy, single-cell RNA sequencing and spatial transcriptomics. We discuss a few representative protocols/assays for each technology along with their strengths, weaknesses, optimal use-cases, and their current stage of clinical deployment by summarizing trial data. We focus on how these technologies help us develop a better understanding of cancer as a rapidly evolving heterogeneous genetic disease that modulates its immediate microenvironment leading to systemic macro-level changes in the patient body. We summarize the review with a flowchart that integrates these three technologies in the existing workflows of clinicians and researchers toward robust detection, accurate diagnosis, and precision oncology.

3D-cardiomics: A spatial transcriptional atlas of the mammalian heart

Understanding the spatial gene expression and regulation in the heart is key to uncovering its developmental and physiological processes, during homeostasis and disease. Numerous techniques exist to gain gene expression and regulation information in organs such as the heart, but few utilize intuitive true-to-life three-dimensional representations to analyze and visualise results. Here we combined transcriptomics with 3D-modelling to interrogate spatial gene expression in the mammalian heart. For this, we microdissected and sequenced transcriptome-wide 18 anatomical sections of the adult mouse heart. Our study has unveiled known and novel genes that display complex spatial expression in the heart sub-compartments. We have also created 3D-cardiomics, an interface for spatial transcriptome analysis and visualization that allows the easy exploration of these data in a 3D model of the heart. 3D-cardiomics is accessible from http://3d-cardiomics.erc.monash.edu/.

Potentials of single-cell genomics in deciphering cellular phenotypes

Single-cell genomics, particularly single-cell transcriptome profiling by RNA sequencing have transformed the possibilities to relate genes to functions, structures, and eventually phenotypes. We can now observe changes in each cell's transcriptome and among its neighborhoods, interrogate the sequence of transcriptional events, and assess their influence on subsequent events. This paradigm shift in biology enables us to infer causal relationships in these events with high accuracy. Here we review the latest single-cell studies in plants that uncover how cellular phenotypes emerge as a result of the transcriptome process such as waves of expression, trajectories of development and responses to the environment, and spatial information. With an eye on the advances made in animal and human studies, we further highlight some of the needed areas for future research and development, including computational methods.

CoSTA: unsupervised convolutional neural network learning for spatial transcriptomics analysis

Background

The rise of spatial transcriptomics technologies is leading to new insights about how gene regulation happens in a spatial context. Determining which genes are expressed in similar spatial patterns can reveal gene regulatory relationships across cell types in a tissue. However, many current analysis methods do not take full advantage of the spatial organization of the data, instead treating pixels as independent features. Here, we present CoSTA: a novel approach to learn spatial similarities between gene expression matrices via convolutional neural network (ConvNet) clustering.

Results

By analyzing simulated and previously published spatial transcriptomics data, we demonstrate that CoSTA learns spatial relationships between genes in a way that emphasizes broader spatial patterns rather than pixel-level correlation. CoSTA provides a quantitative measure of expression pattern similarity between each pair of genes rather than only classifying genes into categories. We find that CoSTA identifies narrower, but biologically relevant, sets of significantly related genes as compared to other approaches.

Conclusions

The deep learning CoSTA approach provides a different angle to spatial transcriptomics analysis by focusing on the shape of expression patterns, using more information about the positions of neighboring pixels than would an overlap or pixel correlation approach. CoSTA can be applied to any spatial transcriptomics data represented in matrix form and may have future applications to datasets such as histology in which images of different genes are from similar but not identical biological sections.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12859-021-04314-1.

The Pandora's box of novel technologies that may revolutionize lung cancer

Non-small cell lung cancer (NSCLC) is one of the most common cancers globally and has a 5-year survival rate ~20%. Immunotherapies have demonstrated long-term and durable responses in NSCLC patients, although they appear to be effective in only a subset of patients. A more comprehensive understanding of the underlying tumour biology may contribute to identifying those patients likely to achieve optimal outcomes. Profiling the tumour microenvironment (TME) has shown to be beneficial in addressing fundamental tumour-immune cell interactions. Advances in multiplexing immunohistochemistry and molecular barcoding has led to recent advances in profiling genes and proteins in NSCLC. Here, we review the recent advancements in spatial profiling technologies for the analysis of NSCLC tissue samples to gain new insights and therapeutic options for NSCLC. The combination of spatial transcriptomics combined with advanced imaging is likely to lead to deep insights into NSCLC tissue biology, which can be a powerful tool to predict likelihood of response to therapy.