Mapping single-cell data to reference atlases by transfer learning

Multi-modal generative modeling for joint analysis of single-cell T cell receptor and gene expression data

Recent advances in single-cell immune profiling have enabled the simultaneous measurement of transcriptome and T cell receptor (TCR) sequences, offering great potential for studying immune responses at the cellular level. However, integrating these diverse modalities across datasets is challenging due to their unique data characteristics and technical variations. Here, to address this, we develop the multimodal generative model mvTCR to fuse modality-specific information across transcriptome and TCR into a shared representation. Our analysis demonstrates the added value of multimodal over unimodal approaches to capture antigen specificity. Notably, we use mvTCR to distinguish T cell subpopulations binding to SARS-CoV-2 antigens from bystander cells. Furthermore, when combined with reference mapping approaches, mvTCR can map newly generated datasets to extensive T cell references, facilitating knowledge transfer. In summary, we envision mvTCR to enable a scalable analysis of multimodal immune profiling data and advance our understanding of immune responses.

Neural Network-Based Filter Design for Compressive Raman Classification of Cells

Cell-based therapies are bound to revolutionize medicine, but significant technical hurdles must be overcome before wider adoption. In particular, nondestructive, label-free methods to characterize cells in real time are needed to optimize the production process and improve quality control. Raman spectroscopy, which provides a fingerprint of a cell's chemical composition, would be an ideal modality but is too slow for high-throughput applications. Compressive Raman techniques, which measure only linear combinations of Raman intensities, can be fast but require careful optimization to deliver high performance. Here, we develop a neural network model to identify optimal parameters for a compressive sensing scheme that reduces measurement time by 2 orders of magnitude. In a data set containing Raman spectra of three different cell types, it achieves up to 90% classification accuracy using only five linear combinations of Raman intensities. Our method thus unlocks the power of Raman spectroscopy for the characterization of cell products.

Pan-cancer profiling of tumor-infiltrating natural killer cells through transcriptional reference mapping

The functional diversity of natural killer (NK) cell repertoires stems from differentiation, homeostatic, receptor-ligand interactions and adaptive-like responses to viral infections. In the present study, we generated a single-cell transcriptional reference map of healthy human blood- and tissue-derived NK cells, with temporal resolution and fate-specific expression of gene-regulatory networks defining NK cell differentiation. Transfer learning facilitated incorporation of tumor-infiltrating NK cell transcriptomes (39 datasets, 7 solid tumors, 427 patients) into the reference map to analyze tumor microenvironment (TME)-induced perturbations. Of the six functionally distinct NK cell states identified, a dysfunctional stressed CD56bright state susceptible to TME-induced immunosuppression and a cytotoxic TME-resistant effector CD56dim state were commonly enriched across tumor types, the ratio of which was predictive of patient outcome in malignant melanoma and osteosarcoma. This resource may inform the design of new NK cell therapies and can be extended through transfer learning to interrogate new datasets from experimental perturbations or disease conditions.

Revealing gene function with statistical inference at single-cell resolution

Single-cell and spatial molecular profiling assays have shown large gains in sensitivity, resolution and throughput. Applying these technologies to specimens from human and model organisms promises to comprehensively catalogue cell types, reveal their lineage origins in development and discern their contributions to disease pathogenesis. Moreover, rapidly dropping costs have made well-controlled perturbation experiments and cohort studies widely accessible, illuminating mechanisms that give rise to phenotypes at the scale of the cell, the tissue and the whole organism. Interpreting the coming flood of single-cell data, much of which will be spatially resolved, will place a tremendous burden on existing computational pipelines. However, statistical concepts, models, tools and algorithms can be repurposed to solve problems now arising in genetic and molecular biology studies of development and disease. Here, I review how the questions that recent technological innovations promise to answer can be addressed by the major classes of statistical tools.

Single-cell Profiling of Reprogrammed Human Neural Stem Cells Unveils High Similarity to Neural Progenitors in the Developing Central Nervous System

BACKGROUND

Similar to induced pluripotent cells (iPSCs), induced neural stem cells (iNSCs) can be directly converted from human somatic cells such as dermal fibroblasts and peripheral blood monocytes. While previous studies have demonstrated the resemblance of iNSCs to neural stem cells derived from primary sources and embryonic stem cells, respectively, a comprehensive analysis of the correlation between iNSCs and their physiological counterparts remained to be investigated.

METHODS

Nowadays, single-cell sequencing technologies provide unique opportunities for in-depth cellular benchmarking of complex cell populations. Our study involves the comprehensive profiling of converted human iNSCs at a single-cell transcriptomic level, alongside conventional methods, like flow cytometry and immunofluorescence stainings.

RESULTS

Our results show that the iNSC conversion yields a homogeneous cell population expressing bona fide neural stem cell markers. Extracting transcriptomic signatures from published single cell transcriptomic atlas data and comparison to the iNSC transcriptome reveals resemblance to embryonic neuroepithelial cells of early neurodevelopmental stages observed in vivo at 5 weeks of development.

CONCLUSION

Our data underscore the physiological relevance of directly converted iNSCs, making them a valuable in vitro system for modeling human central nervous system development and establishing translational applications in cell therapy and compound screening.

SIMS: A deep-learning label transfer tool for single-cell RNA sequencing analysis

Cell atlases serve as vital references for automating cell labeling in new samples, yet existing classification algorithms struggle with accuracy. Here we introduce SIMS (scalable, interpretable machine learning for single cell), a low-code data-efficient pipeline for single-cell RNA classification. We benchmark SIMS against datasets from different tissues and species. We demonstrate SIMS's efficacy in classifying cells in the brain, achieving high accuracy even with small training sets (<3,500 cells) and across different samples. SIMS accurately predicts neuronal subtypes in the developing brain, shedding light on genetic changes during neuronal differentiation and postmitotic fate refinement. Finally, we apply SIMS to single-cell RNA datasets of cortical organoids to predict cell identities and uncover genetic variations between cell lines. SIMS identifies cell-line differences and misannotated cell lineages in human cortical organoids derived from different pluripotent stem cell lines. Altogether, we show that SIMS is a versatile and robust tool for cell-type classification from single-cell datasets.

Leveraging Multi-Tissue, Single-Cell Atlases as Tools to Elucidate Shared Mechanisms of Immune-Mediated Inflammatory Diseases

The observation that certain therapeutic strategies for targeting inflammation benefit patients with distinct immune-mediated inflammatory diseases (IMIDs) is exemplified by the success of TNF blockade in conditions including rheumatoid arthritis, ulcerative colitis, and skin psoriasis, albeit only for subsets of individuals with each condition. This suggests intersecting "nodes" in inflammatory networks at a molecular and cellular level may drive and/or maintain IMIDs, being "shared" between traditionally distinct diagnoses without mapping neatly to a single clinical phenotype. In line with this proposition, integrative tumour tissue analyses in oncology have highlighted novel cell states acting across diverse cancers, with important implications for precision medicine. Drawing upon advances in the oncology field, this narrative review will first summarise learnings from the Human Cell Atlas in health as a platform for interrogating IMID tissues. It will then review cross-disease studies to date that inform this endeavour before considering future directions in the field.

Machine learning integrative approaches to advance computational immunology

The study of immunology, traditionally reliant on proteomics to evaluate individual immune cells, has been revolutionized by single-cell RNA sequencing. Computational immunologists play a crucial role in analysing these datasets, moving beyond traditional protein marker identification to encompass a more detailed view of cellular phenotypes and their functional roles. Recent technological advancements allow the simultaneous measurements of multiple cellular components-transcriptome, proteome, chromatin, epigenetic modifications and metabolites-within single cells, including in spatial contexts within tissues. This has led to the generation of complex multiscale datasets that can include multimodal measurements from the same cells or a mix of paired and unpaired modalities. Modern machine learning (ML) techniques allow for the integration of multiple "omics" data without the need for extensive independent modelling of each modality. This review focuses on recent advancements in ML integrative approaches applied to immunological studies. We highlight the importance of these methods in creating a unified representation of multiscale data collections, particularly for single-cell and spatial profiling technologies. Finally, we discuss the challenges of these holistic approaches and how they will be instrumental in the development of a common coordinate framework for multiscale studies, thereby accelerating research and enabling discoveries in the computational immunology field.

Automated classification of cellular expression in multiplexed imaging data with Nimbus

Multiplexed imaging offers a powerful approach to characterize the spatial topography of tissues in both health and disease. To analyze such data, the specific combination of markers that are present in each cell must be enumerated to enable accurate phenotyping, a process that often relies on unsupervised clustering. We constructed the Pan-Multiplex (Pan-M) dataset containing 197 million distinct annotations of marker expression across 15 different cell types. We used Pan-M to create Nimbus, a deep learning model to predict marker positivity from multiplexed image data. Nimbus is a pre-trained model that uses the underlying images to classify marker expression across distinct cell types, from different tissues, acquired using different microscope platforms, without requiring any retraining. We demonstrate that Nimbus predictions capture the underlying staining patterns of the full diversity of markers present in Pan-M. We then show how Nimbus predictions can be integrated with downstream clustering algorithms to robustly identify cell subtypes in image data. We have open-sourced Nimbus and Pan-M to enable community use at https://github.com/angelolab/Nimbus-Inference.

Deep generative models in single-cell omics

Deep Generative Models (DGMs) are becoming instrumental for inferring probability distributions inherent to complex processes, such as most questions in biomedical research. For many years, there was a lack of mathematical methods that would allow this inference in the scarce data scenario of biomedical research. The advent of single-cell omics has finally made square the so-called "skinny matrix", allowing to apply mathematical methods already extensively used in other areas. Moreover, it is now possible to integrate data at different molecular levels in thousands or even millions of samples, thanks to the number of single-cell atlases being collaboratively generated. Additionally, DGMs have proven useful in other frequent tasks in single-cell analysis pipelines, from dimensionality reduction, cell type annotation to RNA velocity inference. In spite of its promise, DGMs need to be used with caution in biomedical research, paying special attention to its use to answer the right questions and the definition of appropriate error metrics and validation check points that confirm not only its correct use but also its relevance. All in all, DGMs provide an exciting tool that opens a bright future for the integrative analysis of single-cell -omics to understand health and disease.

Single-cell and spatial transcriptomics analysis of non-small cell lung cancer

Lung cancer is the second most frequently diagnosed cancer and the leading cause of cancer-related mortality worldwide. Tumour ecosystems feature diverse immune cell types. Myeloid cells, in particular, are prevalent and have a well-established role in promoting the disease. In our study, we profile approximately 900,000 cells from 25 treatment-naive patients with adenocarcinoma and squamous-cell carcinoma by single-cell and spatial transcriptomics. We note an inverse relationship between anti-inflammatory macrophages and NK cells/T cells, and with reduced NK cell cytotoxicity within the tumour. While we observe a similar cell type composition in both adenocarcinoma and squamous-cell carcinoma, we detect significant differences in the co-expression of various immune checkpoint inhibitors. Moreover, we reveal evidence of a transcriptional "reprogramming" of macrophages in tumours, shifting them towards cholesterol export and adopting a foetal-like transcriptional signature which promotes iron efflux. Our multi-omic resource offers a high-resolution molecular map of tumour-associated macrophages, enhancing our understanding of their role within the tumour microenvironment.

scPLAN: a hierarchical computational framework for single transcriptomics data annotation, integration and cell-type label refinement

MOTIVATION

In the past decade, single-cell RNA sequencing (scRNA-seq) has emerged as a pivotal method for transcriptomic profiling in biomedical research. Precise cell-type identification is crucial for subsequent analysis of single-cell data. And the integration and refinement of annotated data are essential for building comprehensive databases. However, prevailing annotation techniques often overlook the hierarchical organization of cell types, resulting in inconsistent annotations. Meanwhile, most existing integration approaches fail to integrate datasets with different annotation depths and none of them can enhance the labels of outdated data with lower annotation resolutions using more intricately annotated datasets or novel biological findings.

RESULTS

Here, we introduce scPLAN, a hierarchical computational framework designed for scRNA-seq data analysis. scPLAN excels in annotating unlabeled scRNA-seq data using a reference dataset structured along a hierarchical cell-type tree. It identifies potential novel cell types in a systematic, layer-by-layer manner. Additionally, scPLAN effectively integrates annotated scRNA-seq datasets with varying levels of annotation depth, ensuring consistent refinement of cell-type labels across datasets with lower resolutions. Through extensive annotation and novel cell detection experiments, scPLAN has demonstrated its efficacy. Two case studies have been conducted to showcase how scPLAN integrates datasets with diverse cell-type label resolutions and refine their cell-type labels.

AVAILABILITY

https://github.com/michaelGuo1204/scPLAN.

CellWalker2: multi-omic discovery of hierarchical cell type relationships and their associations with genomic annotations

CellWalker2 is a graph diffusion-based method for single-cell genomics data integration. It extends the CellWalker model by incorporating hierarchical relationships between cell types, providing estimates of statistical significance, and adding data structures for analyzing multi-omics data so that gene expression and open chromatin can be jointly modeled. Our open-source software enables users to annotate cells using existing ontologies and to probabilistically match cell types between two or more contexts, including across species. CellWalker2 can also map genomic regions to cell ontologies, enabling precise annotation of elements derived from bulk data, such as enhancers, genetic variants, and sequence motifs. Through simulation studies, we show that CellWalker2 performs better than existing methods in cell type annotation and mapping. We then use data from the brain and immune system to demonstrate CellWalker2's ability to discover cell type-specific regulatory programs and both conserved and divergent cell type relationships in complex tissues.

The future of rapid and automated single-cell data analysis using reference mapping

As the number of single-cell datasets continues to grow rapidly, workflows that map new data to well-curated reference atlases offer enormous promise for the biological community. In this perspective, we discuss key computational challenges and opportunities for single-cell reference-mapping algorithms. We discuss how mapping algorithms will enable the integration of diverse datasets across disease states, molecular modalities, genetic perturbations, and diverse species and will eventually replace manual and laborious unsupervised clustering pipelines.

Data normalization for addressing the challenges in the analysis of single-cell transcriptomic datasets

BACKGROUND

Normalization is a critical step in the analysis of single-cell RNA-sequencing (scRNA-seq) datasets. Its main goal is to make gene counts comparable within and between cells. To do so, normalization methods must account for technical and biological variability. Numerous normalization methods have been developed addressing different sources of dispersion and making specific assumptions about the count data.

MAIN BODY

The selection of a normalization method has a direct impact on downstream analysis, for example differential gene expression and cluster identification. Thus, the objective of this review is to guide the reader in making an informed decision on the most appropriate normalization method to use. To this aim, we first give an overview of the different single cell sequencing platforms and methods commonly used including isolation and library preparation protocols. Next, we discuss the inherent sources of variability of scRNA-seq datasets. We describe the categories of normalization methods and include examples of each. We also delineate imputation and batch-effect correction methods. Furthermore, we describe data-driven metrics commonly used to evaluate the performance of normalization methods. We also discuss common scRNA-seq methods and toolkits used for integrated data analysis.

CONCLUSIONS

According to the correction performed, normalization methods can be broadly classified as within and between-sample algorithms. Moreover, with respect to the mathematical model used, normalization methods can further be classified into: global scaling methods, generalized linear models, mixed methods, and machine learning-based methods. Each of these methods depict pros and cons and make different statistical assumptions. However, there is no better performing normalization method. Instead, metrics such as silhouette width, K-nearest neighbor batch-effect test, or Highly Variable Genes are recommended to assess the performance of normalization methods.

PPPCT: Privacy-Preserving framework for Parallel Clustering Transcriptomics data

Single-cell transcriptomics data provides crucial insights into patients' health, yet poses significant privacy concerns. Genomic data privacy attacks can have deep implications, encompassing not only the patients' health information but also extending widely to compromise their families'. Moreover, the permanence of leaked data exacerbates the challenges, making retraction an impossibility. While extensive efforts have been directed towards clustering single-cell transcriptomics data, addressing critical challenges, especially in the realm of privacy, remains pivotal. This paper introduces an efficient, fast, privacy-preserving approach for clustering single-cell RNA-sequencing (scRNA-seq) datasets. The key contributions include ensuring data privacy, achieving high-quality clustering, accommodating the high dimensionality inherent in the datasets, and maintaining reasonable computation time for big-scale datasets. Our proposed approach utilizes the map-reduce scheme to parallelize clustering, addressing intensive calculation challenges. Intel Software Guard eXtension (SGX) processors are used to ensure the security of sensitive code and data during processing. Additionally, the approach incorporates a logarithm transformation as a preprocessing step, employs non-negative matrix factorization for dimensionality reduction, and utilizes parallel k-means for clustering. The approach fully leverages the computing capabilities of all processing resources within a secure private cloud environment. Experimental results demonstrate the efficacy of our approach in preserving patient privacy while surpassing state-of-the-art methods in both clustering quality and computation time. Our method consistently achieves a minimum of 7% higher Adjusted Rand Index (ARI) than existing approaches, contingent on dataset size. Additionally, due to parallel computations and dimensionality reduction, our approach exhibits efficiency, converging to very good results in less than 10 seconds for a scRNA-seq dataset with 5000 genes and 6000 cells when prioritizing privacy and under two seconds without privacy considerations. Availability and implementation Code and datasets availability: https://github.com/University-of-Windsor/PPPCT.

Understanding glioblastoma at the single-cell level: Recent advances and future challenges

Glioblastoma, the most aggressive and prevalent form of primary brain tumor, is characterized by rapid growth, diffuse infiltration, and resistance to therapies. Intrinsic heterogeneity and cellular plasticity contribute to its rapid progression under therapy; therefore, there is a need to fully understand these tumors at a single-cell level. Over the past decade, single-cell transcriptomics has enabled the molecular characterization of individual cells within glioblastomas, providing previously unattainable insights into the genetic and molecular features that drive tumorigenesis, disease progression, and therapy resistance. However, despite advances in single-cell technologies, challenges such as high costs, complex data analysis and interpretation, and difficulties in translating findings into clinical practice persist. As single-cell technologies are developed further, more insights into the cellular and molecular heterogeneity of glioblastomas are expected, which will help guide the development of personalized and effective therapies, thereby improving prognosis and quality of life for patients.

Integrated Single-Cell Transcriptomic Atlas of Human Kidney Endothelial Cells

Key Points

We created a comprehensive reference atlas of normal human kidney endothelial cells. We confirmed that endothelial cell types in the human kidney were also highly conserved in the mouse kidney.

Background

Kidney endothelial cells are exposed to different microenvironmental conditions that support specific physiologic processes. However, the heterogeneity of human kidney endothelial cells has not yet been systematically described.

Methods

We reprocessed and integrated seven human kidney control single-cell/single-nucleus RNA sequencing datasets of >200,000 kidney cells in the same process.

Results

We identified five major cell types, 29,992 of which were endothelial cells. Endothelial cell reclustering identified seven subgroups that differed in molecular characteristics and physiologic functions. Mapping new data to a normal kidney endothelial cell atlas allows rapid data annotation and analysis. We confirmed that endothelial cell types in the human kidney were also highly conserved in the mouse kidney and identified endothelial marker genes that were conserved in humans and mice, as well as differentially expressed genes between corresponding subpopulations. Furthermore, combined analysis of single-cell transcriptome data with public genome-wide association study data showed a significant enrichment of endothelial cells, especially arterial endothelial cells, in BP heritability. Finally, we identified M1 and M12 from coexpression networks in endothelial cells that may be deeply involved in BP regulation.

Conclusions

We created a comprehensive reference atlas of normal human kidney endothelial cells that provides the molecular foundation for understanding how the identity and function of kidney endothelial cells are altered in disease, aging, and between species. Finally, we provide a publicly accessible online tool to explore the datasets described in this work (https://vascularmap.jiangnan.edu.cn ).

Characterization of Commercially Available Human Primary Alveolar Epithelial Cells

In vitro lung research requires appropriate cell culture models that adequately mimic in vivo structure and function. Previously, researchers extensively used commercially available and easily expandable A549 and NCI-H441 cells, which replicate some but not all features of alveolar epithelial cells. Specifically, these cells are often restricted by terminally altered expression while lacking important alveolar epithelial characteristics. Of late, human primary alveolar epithelial cells (hPAEpCs) have become commercially available but are so far poorly specified. Here, we applied a comprehensive set of technologies to characterize their morphology, surface marker expression, transcriptomic profile, and functional properties. At optimized seeding numbers of 7,500 cells per square centimeter and growth at a gas-liquid interface, hPAEpCs formed regular monolayers with tight junctions and amiloride-sensitive transepithelial ion transport. Electron microscopy revealed lamellar body and microvilli formation characteristic for alveolar type II cells. Protein and single-cell transcriptomic analyses revealed expression of alveolar type I and type II cell markers; yet, transcriptomic data failed to detect NKX2-1, an important transcriptional regulator of alveolar cell differentiation. With increasing passage number, hPAEpCs transdifferentiated toward alveolar-basal intermediates characterized as SFTPC-, KRT8high, and KRT5- cells. In spite of marked changes in the transcriptome as a function of passaging, Uniform Manifold Approximation and Projection plots did not reveal major shifts in cell clusters, and epithelial permeability was unaffected. The present work delineates optimized culture conditions, cellular characteristics, and functional properties of commercially available hPAEpCs. hPAEpCs may provide a useful model system for studies on drug delivery, barrier function, and transepithelial ion transport in vitro.

The covariance environment defines cellular niches for spatial inference

A key challenge of analyzing data from high-resolution spatial profiling technologies is to suitably represent the features of cellular neighborhoods or niches. Here we introduce the covariance environment (COVET), a representation that leverages the gene-gene covariate structure across cells in the niche to capture the multivariate nature of cellular interactions within it. We define a principled optimal transport-based distance metric between COVET niches that scales to millions of cells. Using COVET to encode spatial context, we developed environmental variational inference (ENVI), a conditional variational autoencoder that jointly embeds spatial and single-cell RNA sequencing data into a latent space. ENVI includes two decoders: one to impute gene expression across the spatial modality and a second to project spatial information onto single-cell data. ENVI can confer spatial context to genomics data from single dissociated cells and outperforms alternatives for imputing gene expression on diverse spatial datasets.

Integration of single-cell proteomic datasets through distinctive proteins in cell clusters

The use of mass spectrometry and antibody-based sequencing technologies at the single-cell level has led to an increase in single-cell proteomic datasets. Integrating these datasets is crucial to eliminate the batch effect that often arises due to their limited sequencing molecules. Although methods for horizontally integrating high-dimensional single-cell transcriptomic datasets can also be applied to single-cell proteomic datasets, a specialized approach explicitly tailored for low-dimensional proteomic datasets may enhance the integration process. Here, we introduce SCPRO-HI, an algorithm for the horizontal integration of antibody-based single-cell proteomic datasets. It utilizes a hierarchical cell anchoring technique to match cells based on the similarity of distinctive proteins for constituting cell clusters. A novel variational auto-encoder model is employed for correcting batch effects on the protein abundances, eliminating the need for mapping them into a new domain. Moreover, we propose a technique for extending the algorithm to high-dimensional datasets. The performance of the SCPRO-HI algorithm is evaluated using simulated and real-world single-cell proteomic datasets. The findings demonstrate our algorithm outperforms state-of-the-art methods, achieving a 75% higher silhouette score while preserving HVPs 13% better. Furthermore, the algorithm shows competitive performance in transcriptomic datasets, suggesting potential for integrating high-dimensional mass-spectrometry-based proteomic datasets.

Comparative Analysis of Single-Cell RNA Sequencing Methods with and without Sample Multiplexing

Single-cell RNA sequencing (scRNA-seq) has emerged as a powerful technique for investigating biological heterogeneity at the single-cell level in human systems and model organisms. Recent advances in scRNA-seq have enabled the pooling of cells from multiple samples into single libraries, thereby increasing sample throughput while reducing technical batch effects, library preparation time, and the overall cost. However, a comparative analysis of scRNA-seq methods with and without sample multiplexing is lacking. In this study, we benchmarked methods from two representative platforms: Parse Biosciences (Parse; with sample multiplexing) and 10x Genomics (10x; without sample multiplexing). By using peripheral blood mononuclear cells (PBMCs) obtained from two healthy individuals, we demonstrate that demultiplexed scRNA-seq data obtained from Parse showed similar cell type frequencies compared to 10x data where samples were not multiplexed. Despite relatively lower cell capture affecting library preparation, Parse can detect rare cell types (e.g., plasmablasts and dendritic cells) which is likely due to its relatively higher sensitivity in gene detection. Moreover, a comparative analysis of transcript quantification between the two platforms revealed platform-specific distributions of gene length and GC content. These results offer guidance for researchers in designing high-throughput scRNA-seq studies.

HyGAnno: hybrid graph neural network-based cell type annotation for single-cell ATAC sequencing data

Reliable cell type annotations are crucial for investigating cellular heterogeneity in single-cell omics data. Although various computational approaches have been proposed for single-cell RNA sequencing (scRNA-seq) annotation, high-quality cell labels are still lacking in single-cell sequencing assay for transposase-accessible chromatin (scATAC-seq) data, because of extreme sparsity and inconsistent chromatin accessibility between datasets. Here, we present a novel automated cell annotation method that transfers cell type information from a well-labeled scRNA-seq reference to an unlabeled scATAC-seq target, via a parallel graph neural network, in a semi-supervised manner. Unlike existing methods that utilize only gene expression or gene activity features, HyGAnno leverages genome-wide accessibility peak features to facilitate the training process. In addition, HyGAnno reconstructs a reference-target cell graph to detect cells with low prediction reliability, according to their specific graph connectivity patterns. HyGAnno was assessed across various datasets, showcasing its strengths in precise cell annotation, generating interpretable cell embeddings, robustness to noisy reference data and adaptability to tumor tissues.

scBOL: a universal cell type identification framework for single-cell and spatial transcriptomics data

MOTIVATION

Over the past decade, single-cell transcriptomic technologies have experienced remarkable advancements, enabling the simultaneous profiling of gene expressions across thousands of individual cells. Cell type identification plays an essential role in exploring tissue heterogeneity and characterizing cell state differences. With more and more well-annotated reference data becoming available, massive automatic identification methods have sprung up to simplify the annotation process on unlabeled target data by transferring the cell type knowledge. However, in practice, the target data often include some novel cell types that are not in the reference data. Most existing works usually classify these private cells as one generic 'unassigned' group and learn the features of known and novel cell types in a coupled way. They are susceptible to the potential batch effects and fail to explore the fine-grained semantic knowledge of novel cell types, thus hurting the model's discrimination ability. Additionally, emerging spatial transcriptomic technologies, such as in situ hybridization, sequencing and multiplexed imaging, present a novel challenge to current cell type identification strategies that predominantly neglect spatial organization. Consequently, it is imperative to develop a versatile method that can proficiently annotate single-cell transcriptomics data, encompassing both spatial and non-spatial dimensions.

RESULTS

To address these issues, we propose a new, challenging yet realistic task called universal cell type identification for single-cell and spatial transcriptomics data. In this task, we aim to give semantic labels to target cells from known cell types and cluster labels to those from novel ones. To tackle this problem, instead of designing a suboptimal two-stage approach, we propose an end-to-end algorithm called scBOL from the perspective of Bipartite prototype alignment. Firstly, we identify the mutual nearest clusters in reference and target data as their potential common cell types. On this basis, we mine the cycle-consistent semantic anchor cells to build the intrinsic structure association between two data. Secondly, we design a neighbor-aware prototypical learning paradigm to strengthen the inter-cluster separability and intra-cluster compactness within each data, thereby inspiring the discriminative feature representations. Thirdly, driven by the semantic-aware prototypical learning framework, we can align the known cell types and separate the private cell types from them among reference and target data. Such an algorithm can be seamlessly applied to various data types modeled by different foundation models that can generate the embedding features for cells. Specifically, for non-spatial single-cell transcriptomics data, we use the autoencoder neural network to learn latent low-dimensional cell representations, and for spatial single-cell transcriptomics data, we apply the graph convolution network to capture molecular and spatial similarities of cells jointly. Extensive results on our carefully designed evaluation benchmarks demonstrate the superiority of scBOL over various state-of-the-art cell type identification methods. To our knowledge, we are the pioneers in presenting this pragmatic annotation task, as well as in devising a comprehensive algorithmic framework aimed at resolving this challenge across varied types of single-cell data. Finally, scBOL is implemented in Python using the Pytorch machine-learning library, and it is freely available at https://github.com/aimeeyaoyao/scBOL.

Just how transformative will AI/ML be for immuno-oncology?

Immuno-oncology involves the study of approaches which harness the patient's immune system to fight malignancies. Immuno-oncology, as with every other biomedical and clinical research field as well as clinical operations, is in the midst of technological revolutions, which vastly increase the amount of available data. Recent advances in artificial intelligence and machine learning (AI/ML) have received much attention in terms of their potential to harness available data to improve insights and outcomes in many areas including immuno-oncology. In this review, we discuss important aspects to consider when evaluating the potential impact of AI/ML applications in the clinic. We highlight four clinical/biomedical challenges relevant to immuno-oncology and how they may be able to be addressed by the latest advancements in AI/ML. These challenges include (1) efficiency in clinical workflows, (2) curation of high-quality image data, (3) finding, extracting and synthesizing text knowledge as well as addressing, and (4) small cohort size in immunotherapeutic evaluation cohorts. Finally, we outline how advancements in reinforcement and federated learning, as well as the development of best practices for ethical and unbiased data generation, are likely to drive future innovations.

Starfysh integrates spatial transcriptomic and histologic data to reveal heterogeneous tumor-immune hubs

Spatially resolved gene expression profiling provides insight into tissue organization and cell-cell crosstalk; however, sequencing-based spatial transcriptomics (ST) lacks single-cell resolution. Current ST analysis methods require single-cell RNA sequencing data as a reference for rigorous interpretation of cell states, mostly do not use associated histology images and are not capable of inferring shared neighborhoods across multiple tissues. Here we present Starfysh, a computational toolbox using a deep generative model that incorporates archetypal analysis and any known cell type markers to characterize known or new tissue-specific cell states without a single-cell reference. Starfysh improves the characterization of spatial dynamics in complex tissues using histology images and enables the comparison of niches as spatial hubs across tissues. Integrative analysis of primary estrogen receptor (ER)-positive breast cancer, triple-negative breast cancer (TNBC) and metaplastic breast cancer (MBC) tissues led to the identification of spatial hubs with patient- and disease-specific cell type compositions and revealed metabolic reprogramming shaping immunosuppressive hubs in aggressive MBC.

Spatial transcriptomics reveals prognosis-associated cellular heterogeneity in the papillary thyroid carcinoma microenvironment

BACKGROUND

Papillary thyroid carcinoma (PTC) is the most common malignant endocrine tumour, and its incidence and prevalence are increasing considerably. Cellular heterogeneity in the tumour microenvironment is important for PTC prognosis. Spatial transcriptomics is a powerful technique for cellular heterogeneity study.

METHODS

In conjunction with a clinical pathologist identification method, spatial transcriptomics was employed to characterise the spatial location and RNA profiles of PTC-associated cells within the tissue sections. The spatial RNA-clinical signature genes for each cell type were extracted and applied to outlining the distribution regions of specific cells on the entire section. The cellular heterogeneity of each cell type was further revealed by ContourPlot analysis, monocle analysis, trajectory analysis, ligand-receptor analysis and Gene Ontology enrichment analysis.

RESULTS

The spatial distribution region of tumour cells, typical and atypical follicular cells (FCs and AFCs) and immune cells were accurately and comprehensively identified in all five PTC tissue sections. AFCs were identified as a transitional state between FCs and tumour cells, exhibiting a higher resemblance to the latter. Three tumour foci were shared among all patients out of the 13 observed. Notably, tumour foci No. 2 displayed elevated expression levels of genes associated with lower relapse-free survival in PTC patients. We discovered key ligand-receptor interactions, including LAMB3-ITGA2, FN1-ITGA3 and FN1-SDC4, involved in the transition of PTC cells from FCs to AFCs and eventually to tumour cells. High expression of these patterns correlated with reduced relapse-free survival. In the tumour immune microenvironment, reduced interaction between myeloid-derived TGFB1 and TGFBR1 in tumour focus No. 2 contributed to tumourigenesis and increased heterogeneity. The spatial RNA-clinical analysis method developed here revealed prognosis-associated cellular heterogeneity in the PTC microenvironment.

CONCLUSIONS

The occurrence of tumour foci No. 2 and three enhanced ligand-receptor interactions in the AFC area/tumour foci reduced the relapse-free survival of PTC patients, potentially leading to improved prognostic strategies and targeted therapies for PTC patients.

Integrating single-cell multi-omics and prior biological knowledge for a functional characterization of the immune system

The immune system comprises diverse specialized cell types that cooperate to defend the host against a wide range of pathogenic threats. Recent advancements in single-cell and spatial multi-omics technologies provide rich information about the molecular state of immune cells. Here, we review how the integration of single-cell and spatial multi-omics data with prior knowledge-gathered from decades of detailed biochemical studies-allows us to obtain functional insights, focusing on gene regulatory processes and cell-cell interactions. We present diverse applications in immunology and critically assess underlying assumptions and limitations. Finally, we offer a perspective on the ongoing technological and algorithmic developments that promise to get us closer to a systemic mechanistic understanding of the immune system.

Cell fate decision in erythropoiesis: Insights from multiomics studies

Every second, the body produces 2 million red blood cells through a process called erythropoiesis. Erythropoiesis is hierarchical in that it results from a series of cell fate decisions whereby hematopoietic stem cells progress toward the erythroid lineage. Single-cell transcriptomic and proteomic approaches have revolutionized the way we understand erythropoiesis, revealing it to be a gradual process that underlies a progressive restriction of fate potential driven by quantitative changes in lineage-specifying transcription factors. Despite these major advances, we still know very little about what cell fate decision entails at the molecular level. Novel approaches that simultaneously measure additional properties in single cells, including chromatin accessibility, transcription factor binding, and/or cell surface proteins are being developed at a fast pace, providing the means to exciting new advances in the near future. In this review, we briefly summarize the main findings obtained from single-cell studies of erythropoiesis, highlight outstanding questions, and suggest recent technological advances to address them.

Integration of human organoids single-cell transcriptomic profiles and human genetics repurposes critical cell type-specific drug targets for severe COVID-19

Human organoids recapitulate the cell type diversity and function of their primary organs holding tremendous potentials for basic and translational research. Advances in single-cell RNA sequencing (scRNA-seq) technology and genome-wide association study (GWAS) have accelerated the biological and therapeutic interpretation of trait-relevant cell types or states. Here, we constructed a computational framework to integrate atlas-level organoid scRNA-seq data, GWAS summary statistics, expression quantitative trait loci, and gene-drug interaction data for distinguishing critical cell populations and drug targets relevant to coronavirus disease 2019 (COVID-19) severity. We found that 39 cell types across eight kinds of organoids were significantly associated with COVID-19 outcomes. Notably, subset of lung mesenchymal stem cells increased proximity with fibroblasts predisposed to repair COVID-19-damaged lung tissue. Brain endothelial cell subset exhibited significant associations with severe COVID-19, and this cell subset showed a notable increase in cell-to-cell interactions with other brain cell types, including microglia. We repurposed 33 druggable genes, including IFNAR2, TYK2, and VIPR2, and their interacting drugs for COVID-19 in a cell-type-specific manner. Overall, our results showcase that host genetic determinants have cellular-specific contribution to COVID-19 severity, and identification of cell type-specific drug targets may facilitate to develop effective therapeutics for treating severe COVID-19 and its complications.

Single-Cell Transcriptomics Reveals Pre-existing COVID-19 Vulnerability Factors in Lung Cancer Patients

Coronavirus disease 2019 (COVID-19) and cancer are major health threats, and individuals may develop both simultaneously. Recent studies have indicated that patients with cancer are particularly vulnerable to COVID-19, but the molecular mechanisms underlying the associations remain poorly understood. To address this knowledge gap, we collected single-cell RNA-sequencing data from COVID-19, lung adenocarcinoma, small cell lung carcinoma patients, and normal lungs to perform an integrated analysis. We characterized altered cell populations, gene expression, and dysregulated intercellular communication in diseases. Our analysis identified pathologic conditions shared by COVID-19 and lung cancer, including upregulated TMPRSS2 expression in epithelial cells, stronger inflammatory responses mediated by macrophages, increased T-cell response suppression, and elevated fibrosis risk by pathologic fibroblasts. These pre-existing conditions in patients with lung cancer may lead to more severe inflammation, fibrosis, and weakened adaptive immune response upon COVID-19 infection. Our findings revealed potential molecular mechanisms driving an increased COVID-19 risk in patients with lung cancer and suggested preventive and therapeutic targets for COVID-19 in this population.

IMPLICATIONS

Our work reveals the potential molecular mechanisms contributing to the vulnerability to COVID-19 in patients with lung cancer.

Spatially organized cellular communities form the developing human heart

The heart, which is the first organ to develop, is highly dependent on its form to function1,2. However, how diverse cardiac cell types spatially coordinate to create the complex morphological structures that are crucial for heart function remains unclear. Here we integrated single-cell RNA-sequencing with high-resolution multiplexed error-robust fluorescence in situ hybridization to resolve the identity of the cardiac cell types that develop the human heart. This approach also provided a spatial mapping of individual cells that enables illumination of their organization into cellular communities that form distinct cardiac structures. We discovered that many of these cardiac cell types further specified into subpopulations exclusive to specific communities, which support their specialization according to the cellular ecosystem and anatomical region. In particular, ventricular cardiomyocyte subpopulations displayed an unexpected complex laminar organization across the ventricular wall and formed, with other cell subpopulations, several cellular communities. Interrogating cell-cell interactions within these communities using in vivo conditional genetic mouse models and in vitro human pluripotent stem cell systems revealed multicellular signalling pathways that orchestrate the spatial organization of cardiac cell subpopulations during ventricular wall morphogenesis. These detailed findings into the cellular social interactions and specialization of cardiac cell types constructing and remodelling the human heart offer new insights into structural heart diseases and the engineering of complex multicellular tissues for human heart repair.

scAuto as a comprehensive framework for single-cell chromatin accessibility data analysis

Interpreting single-cell chromatin accessibility data is crucial for understanding intercellular heterogeneity regulation. Despite the progress in computational methods for analyzing this data, there is still a lack of a comprehensive analytical framework and a user-friendly online analysis tool. To fill this gap, we developed a pre-trained deep learning-based framework, single-cell auto-correlation transformers (scAuto), to overcome the challenge. Following DNABERT's methodology of pre-training and fine-tuning, scAuto learns a general understanding of DNA sequence's grammar by being pre-trained on unlabeled human genome via self-supervision; it is then transferred to the single-cell chromatin accessibility analysis task of scATAC-seq data for supervised fine-tuning. We extensively validated scAuto on the Buenrostro2018 dataset, demonstrating its superior performance on chromatin accessibility prediction, single-cell clustering, and data denoising. Based on scAuto, we further developed an interactive web server for single-cell chromatin accessibility data analysis. It integrates tutorial-style interfaces for those with limited programming skills. The platform is accessible at http://zhanglab.icaup.cn. To our knowledge, this work is expected to help analyze single-cell chromatin accessibility data and facilitate the development of precision medicine.

scGPT: toward building a foundation model for single-cell multi-omics using generative AI

Generative pretrained models have achieved remarkable success in various domains such as language and computer vision. Specifically, the combination of large-scale diverse datasets and pretrained transformers has emerged as a promising approach for developing foundation models. Drawing parallels between language and cellular biology (in which texts comprise words; similarly, cells are defined by genes), our study probes the applicability of foundation models to advance cellular biology and genetic research. Using burgeoning single-cell sequencing data, we have constructed a foundation model for single-cell biology, scGPT, based on a generative pretrained transformer across a repository of over 33 million cells. Our findings illustrate that scGPT effectively distills critical biological insights concerning genes and cells. Through further adaptation of transfer learning, scGPT can be optimized to achieve superior performance across diverse downstream applications. This includes tasks such as cell type annotation, multi-batch integration, multi-omic integration, perturbation response prediction and gene network inference.

A single cell atlas of frozen shoulder capsule identifies features associated with inflammatory fibrosis resolution

Frozen shoulder is a spontaneously self-resolving chronic inflammatory fibrotic human disease, which distinguishes the condition from most fibrotic diseases that are progressive and irreversible. Using single-cell analysis, we identify pro-inflammatory MERTKlowCD48+ macrophages and MERTK + LYVE1 + MRC1+ macrophages enriched for negative regulators of inflammation which co-exist in frozen shoulder capsule tissues. Micro-cultures of patient-derived cells identify integrin-mediated cell-matrix interactions between MERTK+ macrophages and pro-resolving DKK3+ and POSTN+ fibroblasts, suggesting that matrix remodelling plays a role in frozen shoulder resolution. Cross-tissue analysis reveals a shared gene expression cassette between shoulder capsule MERTK+ macrophages and a respective population enriched in synovial tissues of rheumatoid arthritis patients in disease remission, supporting the concept that MERTK+ macrophages mediate resolution of inflammation and fibrosis. Single-cell transcriptomic profiling and spatial analysis of human foetal shoulder tissues identify MERTK + LYVE1 + MRC1+ macrophages and DKK3+ and POSTN+ fibroblast populations analogous to those in frozen shoulder, suggesting that the template to resolve fibrosis is established during shoulder development. Crosstalk between MerTK+ macrophages and pro-resolving DKK3+ and POSTN+ fibroblasts could facilitate resolution of frozen shoulder, providing a basis for potential therapeutic resolution of persistent fibrotic diseases.

An atlas of cells in the human tonsil

Palatine tonsils are secondary lymphoid organs (SLOs) representing the first line of immunological defense against inhaled or ingested pathogens. We generated an atlas of the human tonsil composed of >556,000 cells profiled across five different data modalities, including single-cell transcriptome, epigenome, proteome, and immune repertoire sequencing, as well as spatial transcriptomics. This census identified 121 cell types and states, defined developmental trajectories, and enabled an understanding of the functional units of the tonsil. Exemplarily, we stratified myeloid slan-like subtypes, established a BCL6 enhancer as locally active in follicle-associated T and B cells, and identified SIX5 as putative transcriptional regulator of plasma cell maturation. Analyses of a validation cohort confirmed the presence, annotation, and markers of tonsillar cell types and provided evidence of age-related compositional shifts. We demonstrate the value of this resource by annotating cells from B cell-derived mantle cell lymphomas, linking transcriptional heterogeneity to normal B cell differentiation states of the human tonsil.

Molecularly stratified hypothalamic astrocytes are cellular foci for obesity

Astrocytes safeguard the homeostasis of the central nervous system1,2. Despite their prominent morphological plasticity under conditions that challenge the brain's adaptive capacity3-5, the classification of astrocytes, and relating their molecular make-up to spatially devolved neuronal operations that specify behavior or metabolism, remained mostly futile6,7. Although it seems unexpected in the era of single-cell biology, the lack of a major advance in stratifying astrocytes under physiological conditions rests on the incompatibility of 'neurocentric' algorithms that rely on stable developmental endpoints, lifelong transcriptional, neurotransmitter, and neuropeptide signatures for classification6-8 with the dynamic functional states, anatomic allocation, and allostatic plasticity of astrocytes1. Simplistically, therefore, astrocytes are still grouped as 'resting' vs. 'reactive', the latter referring to pathological states marked by various inducible genes3,9,10. Here, we introduced a machine learning-based feature recognition algorithm that benefits from the cumulative power of published single-cell RNA-seq data on astrocytes as a reference map to stepwise eliminate pleiotropic and inducible cellular features. For the healthy hypothalamus, this walk-back approach revealed gene regulatory networks (GRNs) that specified subsets of astrocytes, and could be used as landmarking tools for their anatomical assignment. The core molecular censuses retained by astrocyte subsets were sufficient to stratify them by allostatic competence, chiefly their signaling and metabolic interplay with neurons. Particularly, we found differentially expressed mitochondrial genes in insulin-sensing astrocytes and demonstrated their reciprocal signaling with neurons that work antagonistically within the food intake circuitry. As a proof-of-concept, we showed that disrupting Mfn2 expression in astrocytes reduced their ability to support dynamic circuit reorganization, a time-locked feature of satiety in the hypothalamus, thus leading to obesity in mice. Overall, our results suggest that astrocytes in the healthy brain are fundamentally more heterogeneous than previously thought and topologically mirror the specificity of local neurocircuits.

The impacts of active and self-supervised learning on efficient annotation of single-cell expression data

A crucial step in the analysis of single-cell data is annotating cells to cell types and states. While a myriad of approaches has been proposed, manual labeling of cells to create training datasets remains tedious and time-consuming. In the field of machine learning, active and self-supervised learning methods have been proposed to improve the performance of a classifier while reducing both annotation time and label budget. However, the benefits of such strategies for single-cell annotation have yet to be evaluated in realistic settings. Here, we perform a comprehensive benchmarking of active and self-supervised labeling strategies across a range of single-cell technologies and cell type annotation algorithms. We quantify the benefits of active learning and self-supervised strategies in the presence of cell type imbalance and variable similarity. We introduce adaptive reweighting, a heuristic procedure tailored to single-cell data-including a marker-aware version-that shows competitive performance with existing approaches. In addition, we demonstrate that having prior knowledge of cell type markers improves annotation accuracy. Finally, we summarize our findings into a set of recommendations for those implementing cell type annotation procedures or platforms. An R package implementing the heuristic approaches introduced in this work may be found at https://github.com/camlab-bioml/leader .

Single-cell transcriptomics reveals that glial cells integrate homeostatic and circadian processes to drive sleep-wake cycles

The sleep-wake cycle is determined by circadian and sleep homeostatic processes. However, the molecular impact of these processes and their interaction in different brain cell populations are unknown. To fill this gap, we profiled the single-cell transcriptome of adult Drosophila brains across the sleep-wake cycle and four circadian times. We show cell type-specific transcriptomic changes, with glia displaying the largest variation. Glia are also among the few cell types whose gene expression correlates with both sleep homeostat and circadian clock. The sleep-wake cycle and sleep drive level affect the expression of clock gene regulators in glia, and disrupting clock genes specifically in glia impairs homeostatic sleep rebound after sleep deprivation. These findings provide a comprehensive view of the effects of sleep homeostatic and circadian processes on distinct cell types in an entire animal brain and reveal glia as an interaction site of these two processes to determine sleep-wake dynamics.

A time- and single-cell-resolved model of murine bone marrow hematopoiesis

The paradigmatic hematopoietic tree model is increasingly recognized to be limited, as it is based on heterogeneous populations largely defined by non-homeostatic assays testing cell fate potentials. Here, we combine persistent labeling with time-series single-cell RNA sequencing to build a real-time, quantitative model of in vivo tissue dynamics for murine bone marrow hematopoiesis. We couple cascading single-cell expression patterns with dynamic changes in differentiation and growth speeds. The resulting explicit linkage between molecular states and cellular behavior reveals widely varying self-renewal and differentiation properties across distinct lineages. Transplanted stem cells show strong acceleration of differentiation at specific stages of erythroid and neutrophil production, illustrating how the model can quantify the impact of perturbations. Our reconstruction of dynamic behavior from snapshot measurements is akin to how a kinetoscope allows sequential images to merge into a movie. We posit that this approach is generally applicable to understanding tissue-scale dynamics at high resolution.

Dictionary learning for integrative, multimodal and scalable single-cell analysis

Mapping single-cell sequencing profiles to comprehensive reference datasets provides a powerful alternative to unsupervised analysis. However, most reference datasets are constructed from single-cell RNA-sequencing data and cannot be used to annotate datasets that do not measure gene expression. Here we introduce 'bridge integration', a method to integrate single-cell datasets across modalities using a multiomic dataset as a molecular bridge. Each cell in the multiomic dataset constitutes an element in a 'dictionary', which is used to reconstruct unimodal datasets and transform them into a shared space. Our procedure accurately integrates transcriptomic data with independent single-cell measurements of chromatin accessibility, histone modifications, DNA methylation and protein levels. Moreover, we demonstrate how dictionary learning can be combined with sketching techniques to improve computational scalability and harmonize 8.6 million human immune cell profiles from sequencing and mass cytometry experiments. Our approach, implemented in version 5 of our Seurat toolkit ( http://www.satijalab.org/seurat ), broadens the utility of single-cell reference datasets and facilitates comparisons across diverse molecular modalities.

Representing and extracting knowledge from single-cell data

Single-cell analysis is currently one of the most high-resolution techniques to study biology. The large complex datasets that have been generated have spurred numerous developments in computational biology, in particular the use of advanced statistics and machine learning. This review attempts to explain the deeper theoretical concepts that underpin current state-of-the-art analysis methods. Single-cell analysis is covered from cell, through instruments, to current and upcoming models. The aim of this review is to spread concepts which are not yet in common use, especially from topology and generative processes, and how new statistical models can be developed to capture more of biology. This opens epistemological questions regarding our ontology and models, and some pointers will be given to how natural language processing (NLP) may help overcome our cognitive limitations for understanding single-cell data.

Semi-supervised integration of single-cell transcriptomics data

Batch effects in single-cell RNA-seq data pose a significant challenge for comparative analyses across samples, individuals, and conditions. Although batch effect correction methods are routinely applied, data integration often leads to overcorrection and can result in the loss of biological variability. In this work we present STACAS, a batch correction method for scRNA-seq that leverages prior knowledge on cell types to preserve biological variability upon integration. Through an open-source benchmark, we show that semi-supervised STACAS outperforms state-of-the-art unsupervised methods, as well as supervised methods such as scANVI and scGen. STACAS scales well to large datasets and is robust to incomplete and imprecise input cell type labels, which are commonly encountered in real-life integration tasks. We argue that the incorporation of prior cell type information should be a common practice in single-cell data integration, and we provide a flexible framework for semi-supervised batch effect correction.

SiFT: uncovering hidden biological processes by probabilistic filtering of single-cell data

Cellular populations simultaneously encode multiple biological attributes, including spatial configuration, temporal trajectories, and cell-cell interactions. Some of these signals may be overshadowed by others and harder to recover, despite the great progress made to computationally reconstruct biological processes from single-cell data. To address this, we present SiFT, a kernel-based projection method for filtering biological signals in single-cell data, thus uncovering underlying biological processes. SiFT applies to a wide range of tasks, from the removal of unwanted variation in the data to revealing hidden biological structures. We demonstrate how SiFT enhances the liver circadian signal by filtering spatial zonation, recovers regenerative cell subpopulations in spatially-resolved liver data, and exposes COVID-19 disease-related cells, pathways, and dynamics by filtering healthy reference signals. SiFT performs the correction at the gene expression level, can scale to large datasets, and compares favorably to state-of-the-art methods.

InClust+: the deep generative framework with mask modules for multimodal data integration, imputation, and cross-modal generation

BACKGROUND

With the development of single-cell technology, many cell traits can be measured. Furthermore, the multi-omics profiling technology could jointly measure two or more traits in a single cell simultaneously. In order to process the various data accumulated rapidly, computational methods for multimodal data integration are needed.

RESULTS

Here, we present inClust+, a deep generative framework for the multi-omics. It's built on previous inClust that is specific for transcriptome data, and augmented with two mask modules designed for multimodal data processing: an input-mask module in front of the encoder and an output-mask module behind the decoder. InClust+ was first used to integrate scRNA-seq and MERFISH data from similar cell populations, and to impute MERFISH data based on scRNA-seq data. Then, inClust+ was shown to have the capability to integrate the multimodal data (e.g. tri-modal data with gene expression, chromatin accessibility and protein abundance) with batch effect. Finally, inClust+ was used to integrate an unlabeled monomodal scRNA-seq dataset and two labeled multimodal CITE-seq datasets, transfer labels from CITE-seq datasets to scRNA-seq dataset, and generate the missing modality of protein abundance in monomodal scRNA-seq data. In the above examples, the performance of inClust+ is better than or comparable to the most recent tools in the corresponding task.

CONCLUSIONS

The inClust+ is a suitable framework for handling multimodal data. Meanwhile, the successful implementation of mask in inClust+ means that it can be applied to other deep learning methods with similar encoder-decoder architecture to broaden the application scope of these models.

Mosaic integration and knowledge transfer of single-cell multimodal data with MIDAS

Integrating single-cell datasets produced by multiple omics technologies is essential for defining cellular heterogeneity. Mosaic integration, in which different datasets share only some of the measured modalities, poses major challenges, particularly regarding modality alignment and batch effect removal. Here, we present a deep probabilistic framework for the mosaic integration and knowledge transfer (MIDAS) of single-cell multimodal data. MIDAS simultaneously achieves dimensionality reduction, imputation and batch correction of mosaic data by using self-supervised modality alignment and information-theoretic latent disentanglement. We demonstrate its superiority to 19 other methods and reliability by evaluating its performance in trimodal and mosaic integration tasks. We also constructed a single-cell trimodal atlas of human peripheral blood mononuclear cells and tailored transfer learning and reciprocal reference mapping schemes to enable flexible and accurate knowledge transfer from the atlas to new data. Applications in mosaic integration, pseudotime analysis and cross-tissue knowledge transfer on bone marrow mosaic datasets demonstrate the versatility and superiority of MIDAS. MIDAS is available at https://github.com/labomics/midas .

Continually adapting pre-trained language model to universal annotation of single-cell RNA-seq data

MOTIVATION

Cell-type annotation of single-cell RNA-sequencing (scRNA-seq) data is a hallmark of biomedical research and clinical application. Current annotation tools usually assume the simultaneous acquisition of well-annotated data, but without the ability to expand knowledge from new data. Yet, such tools are inconsistent with the continuous emergence of scRNA-seq data, calling for a continuous cell-type annotation model. In addition, by their powerful ability of information integration and model interpretability, transformer-based pre-trained language models have led to breakthroughs in single-cell biology research. Therefore, the systematic combining of continual learning and pre-trained language models for cell-type annotation tasks is inevitable.

RESULTS

We herein propose a universal cell-type annotation tool, called CANAL, that continuously fine-tunes a pre-trained language model trained on a large amount of unlabeled scRNA-seq data, as new well-labeled data emerges. CANAL essentially alleviates the dilemma of catastrophic forgetting, both in terms of model inputs and outputs. For model inputs, we introduce an experience replay schema that repeatedly reviews previous vital examples in current training stages. This is achieved through a dynamic example bank with a fixed buffer size. The example bank is class-balanced and proficient in retaining cell-type-specific information, particularly facilitating the consolidation of patterns associated with rare cell types. For model outputs, we utilize representation knowledge distillation to regularize the divergence between previous and current models, resulting in the preservation of knowledge learned from past training stages. Moreover, our universal annotation framework considers the inclusion of new cell types throughout the fine-tuning and testing stages. We can continuously expand the cell-type annotation library by absorbing new cell types from newly arrived, well-annotated training datasets, as well as automatically identify novel cells in unlabeled datasets. Comprehensive experiments with data streams under various biological scenarios demonstrate the versatility and high model interpretability of CANAL.

AVAILABILITY

An implementation of CANAL is available from https://github.com/aster-ww/CANAL-torch.

CONTACT

dengmh@pku.edu.cn.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Journal Name online.

Deep learning in spatially resolved transcriptfomics: a comprehensive technical view

Spatially resolved transcriptomics (SRT) is a pioneering method for simultaneously studying morphological contexts and gene expression at single-cell precision. Data emerging from SRT are multifaceted, presenting researchers with intricate gene expression matrices, precise spatial details and comprehensive histology visuals. Such rich and intricate datasets, unfortunately, render many conventional methods like traditional machine learning and statistical models ineffective. The unique challenges posed by the specialized nature of SRT data have led the scientific community to explore more sophisticated analytical avenues. Recent trends indicate an increasing reliance on deep learning algorithms, especially in areas such as spatial clustering, identification of spatially variable genes and data alignment tasks. In this manuscript, we provide a rigorous critique of these advanced deep learning methodologies, probing into their merits, limitations and avenues for further refinement. Our in-depth analysis underscores that while the recent innovations in deep learning tailored for SRT have been promising, there remains a substantial potential for enhancement. A crucial area that demands attention is the development of models that can incorporate intricate biological nuances, such as phylogeny-aware processing or in-depth analysis of minuscule histology image segments. Furthermore, addressing challenges like the elimination of batch effects, perfecting data normalization techniques and countering the overdispersion and zero inflation patterns seen in gene expression is pivotal. To support the broader scientific community in their SRT endeavors, we have meticulously assembled a comprehensive directory of readily accessible SRT databases, hoping to serve as a foundation for future research initiatives.

scEVOLVE: cell-type incremental annotation without forgetting for single-cell RNA-seq data

The evolution in single-cell RNA sequencing (scRNA-seq) technology has opened a new avenue for researchers to inspect cellular heterogeneity with single-cell precision. One crucial aspect of this technology is cell-type annotation, which is fundamental for any subsequent analysis in single-cell data mining. Recently, the scientific community has seen a surge in the development of automatic annotation methods aimed at this task. However, these methods generally operate at a steady-state total cell-type capacity, significantly restricting the cell annotation systems'capacity for continuous knowledge acquisition. Furthermore, creating a unified scRNA-seq annotation system remains challenged by the need to progressively expand its understanding of ever-increasing cell-type concepts derived from a continuous data stream. In response to these challenges, this paper presents a novel and challenging setting for annotation, namely cell-type incremental annotation. This concept is designed to perpetually enhance cell-type knowledge, gleaned from continuously incoming data. This task encounters difficulty with data stream samples that can only be observed once, leading to catastrophic forgetting. To address this problem, we introduce our breakthrough methodology termed scEVOLVE, an incremental annotation method. This innovative approach is built upon the methodology of contrastive sample replay combined with the fundamental principle of partition confidence maximization. Specifically, we initially retain and replay sections of the old data in each subsequent training phase, then establish a unique prototypical learning objective to mitigate the cell-type imbalance problem, as an alternative to using cross-entropy. To effectively emulate a model that trains concurrently with complete data, we introduce a cell-type decorrelation strategy that efficiently scatters feature representations of each cell type uniformly. We constructed the scEVOLVE framework with simplicity and ease of integration into most deep softmax-based single-cell annotation methods. Thorough experiments conducted on a range of meticulously constructed benchmarks consistently prove that our methodology can incrementally learn numerous cell types over an extended period, outperforming other strategies that fail quickly. As far as our knowledge extends, this is the first attempt to propose and formulate an end-to-end algorithm framework to address this new, practical task. Additionally, scEVOLVE, coded in Python using the Pytorch machine-learning library, is freely accessible at https://github.com/aimeeyaoyao/scEVOLVE.

scATAnno: Automated Cell Type Annotation for single-cell ATAC Sequencing Data

The recent advances in single-cell epigenomic techniques have created a growing demand for scATAC-seq analysis. One key task is to determine cell types based on epigenetic profiling. We introduce scATAnno, a workflow designed to automatically annotate scATAC-seq data using large-scale scATAC-seq reference atlases. This workflow can generate scATAC-seq reference atlases from publicly available datasets, and enable accurate cell type annotation by integrating query data with reference atlases, without the aid of scRNA-seq profiling. To enhance annotation accuracy, we have incorporated KNN-based and weighted distance-based uncertainty scores to effectively detect unknown cell populations within the query data. We showcase the utility of scATAnno across multiple datasets, including peripheral blood mononuclear cell (PBMC), basal cell carcinoma (BCC) and Triple Negative Breast Cancer (TNBC), and demonstrate that scATAnno accurately annotates cell types across conditions. Overall, scATAnno is a powerful tool for cell type annotation in scATAC-seq data and can aid in the interpretation of new scATAC-seq datasets in complex biological systems.