A test metric for assessing single-cell RNA-seq batch correction

scMODAL: a general deep learning framework for comprehensive single-cell multi-omics data alignment with feature links

Recent advancements in single-cell technologies have enabled comprehensive characterization of cellular states through transcriptomic, epigenomic, and proteomic profiling at single-cell resolution. These technologies have significantly deepened our understanding of cell functions and disease mechanisms from various omics perspectives. As these technologies evolve rapidly and data resources expand, there is a growing need for computational methods that can integrate information from different modalities to facilitate joint analysis of single-cell multi-omics data. However, integrating single-cell omics datasets presents unique challenges due to varied feature correlations and technology-specific limitations. To address these challenges, we introduce scMODAL, a deep learning framework tailored for single-cell multi-omics data alignment using feature links. scMODAL integrates datasets with limited known positively correlated features, leveraging neural networks and generative adversarial networks to align cell embeddings and preserve feature topology. Our experiments demonstrate scMODAL's effectiveness in removing unwanted variation, preserving biological information, and accurately identifying cell subpopulations across diverse datasets. scMODAL not only advances integration tasks but also supports downstream analyses such as feature imputation and feature relationship inference, offering a robust solution for advancing single-cell multi-omics research.

A CRISPR/Cas9-based enhancement of high-throughput single-cell transcriptomics

Single-cell RNA-seq (scRNAseq) struggles to capture the cellular heterogeneity of transcripts within individual cells due to the prevalence of highly abundant and ubiquitous transcripts, which can obscure the detection of biologically distinct transcripts expressed up to several orders of magnitude lower levels. To address this challenge, here we introduce single-cell CRISPRclean (scCLEAN), a molecular method that globally recomposes scRNAseq libraries, providing a benefit that cannot be recapitulated with deeper sequencing. scCLEAN utilizes the programmability of CRISPR/Cas9 to target and remove less than 1% of the transcriptome while redistributing approximately half of reads, shifting the focus toward less abundant transcripts. We experimentally apply scCLEAN to both heterogeneous immune cells and homogenous vascular smooth muscle cells to demonstrate its ability to uncover biological signatures in different biological contexts. We further emphasize scCLEAN's versatility by applying it to a third-generation sequencing method, single-cell MAS-Seq, to increase transcript-level detection and discovery. Here we show the possible utility of scCLEAN across a wide array of human tissues and cell types, indicating which contexts this technology proves beneficial and those in which its application is not advisable.

48-Hour and 24-Hour Time-lapse Single-nucleus Transcriptomics Reveal Cell-type specific Circadian Rhythms in Arabidopsis

Functional circadian clock is critical to the adaptation and survival of organisms. In land plants, the comprehensive profiling of circadian gene expression at the single-cell level is largely unknown partly due to the challenges in obtaining precisely-timed single cells embedded within cell walls. To bridge this gap, we employ time-lapse single-nucleus RNA sequencing (snRNA-seq) on Arabidopsis seedlings collected over a 48-hour window at 4-hour intervals, as well as over a 24-hour day at 2-hour intervals, yielding a total of over 77,142 and 130,000 nuclei. Here, we find that four cell clusters in the shoot share a coherent rhythm, while around 3000 genes display cell-type specific rhythmic expression. Our analysis indicates that genes encoding circadian regulators oscillate in multiple cell types, and the majority of them are well-documented core clock genes, suggesting the snRNA-seq circadian data could be used to identify more clock components oscillating in a cell-autonomous way. We identify ABF1 as a circadian regulator, whose overexpression and shortens the circadian period. Our data provides a comprehensive resource for plant circadian rhythmicity at the single-cell level (hosted at https://zhailab.bio.sustech.edu.cn/sc_circadian ).

An integrated transcriptomic cell atlas of human endoderm-derived organoids

Human pluripotent stem cells and tissue-resident fetal and adult stem cells can generate epithelial tissues of endodermal origin in vitro that recapitulate aspects of developing and adult human physiology. Here, we integrate single-cell transcriptomes from 218 samples covering organoids and other models of diverse endoderm-derived tissues to establish an initial version of a human endoderm-derived organoid cell atlas. The integration includes nearly one million cells across diverse conditions, data sources and protocols. We compare cell types and states between organoid models and harmonize cell annotations through mapping to primary tissue counterparts. Focusing on the intestine and lung, we provide examples of mapping data from new protocols and show how the atlas can be used as a diverse cohort to assess perturbations and disease models. The human endoderm-derived organoid cell atlas makes diverse datasets centrally available and will be valuable to assess fidelity, characterize perturbed and diseased states, and streamline protocol development.

Advances and applications in single-cell and spatial genomics

The applications of single-cell and spatial technologies in recent times have revolutionized the present understanding of cellular states and the cellular heterogeneity inherent in complex biological systems. These advancements offer unprecedented resolution in the examination of the functional genomics of individual cells and their spatial context within tissues. In this review, we have comprehensively discussed the historical development and recent progress in the field of single-cell and spatial genomics. We have reviewed the breakthroughs in single-cell multi-omics technologies, spatial genomics methods, and the computational strategies employed toward the analyses of single-cell atlas data. Furthermore, we have highlighted the advances made in constructing cellular atlases and their clinical applications, particularly in the context of disease. Finally, we have discussed the emerging trends, challenges, and opportunities in this rapidly evolving field.

scPRINT: pre-training on 50 million cells allows robust gene network predictions

A cell is governed by the interaction of myriads of macromolecules. Inferring such a network of interactions has remained an elusive milestone in cellular biology. Building on recent advances in large foundation models and their ability to learn without supervision, we present scPRINT, a large cell model for the inference of gene networks pre-trained on more than 50 million cells from the cellxgene database. Using innovative pretraining tasks and model architecture, scPRINT pushes large transformer models towards more interpretability and usability when uncovering the complex biology of the cell. Based on our atlas-level benchmarks, scPRINT demonstrates superior performance in gene network inference to the state of the art, as well as competitive zero-shot abilities in denoising, batch effect correction, and cell label prediction. On an atlas of benign prostatic hyperplasia, scPRINT highlights the profound connections between ion exchange, senescence, and chronic inflammation.

Revealing a coherent cell state landscape across single cell datasets with CONCORD

Batch integration, denoising, and dimensionality reduction remain fundamental challenges in single-cell data analysis. While many machine learning tools aim to overcome these challenges by engineering model architectures, we use a different strategy, building on the insight that optimized mini-batch sampling during training can profoundly influence learning outcomes. We present CONCORD, a self-supervised learning approach that implements a unified, probabilistic data sampling scheme combining neighborhood-aware and dataset-aware sampling: the former enhancing resolution while the latter removing batch effects. Using only a minimalist one-hidden-layer neural network and contrastive learning, CONCORD achieves state-of-the-art performance without relying on deep architectures, auxiliary losses, or supervision. It generates high-resolution cell atlases that seamlessly integrate data across batches, technologies, and species, without relying on prior assumptions about data structure. The resulting latent representations are denoised, interpretable, and biologically meaningful-capturing gene co-expression programs, resolving subtle cellular states, and preserving both local geometric relationships and global topological organization. We demonstrate CONCORD's broad applicability across diverse datasets, establishing it as a general-purpose framework for learning unified, high-fidelity representations of cellular identity and dynamics.

Feature selection methods affect the performance of scRNA-seq data integration and querying

The availability of single-cell transcriptomics has allowed the construction of reference cell atlases, but their usefulness depends on the quality of dataset integration and the ability to map new samples. Previous benchmarks have compared integration methods and suggest that feature selection improves performance but have not explored how best to select features. Here, we benchmark feature selection methods for single-cell RNA sequencing integration using metrics beyond batch correction and preservation of biological variation to assess query mapping, label transfer and the detection of unseen populations. We reinforce common practice by showing that highly variable feature selection is effective for producing high-quality integrations and provide further guidance on the effect of the number of features selected, batch-aware feature selection, lineage-specific feature selection and integration and the interaction between feature selection and integration models. These results are informative for analysts working on large-scale tissue atlases, using atlases or integrating their own data to tackle specific biological questions.

Reference-informed evaluation of batch correction for single-cell omics data with overcorrection awareness

Batch effect correction (BEC) is fundamental to integrate multiple single-cell RNA sequencing datasets, and its success is critical to empower in-depth interrogation for biological insights. However, no simple metric is available to evaluate BEC performance with sensitivity to data overcorrection, which erases true biological variations and leads to false biological discoveries. Here, we propose RBET, a reference-informed statistical framework for evaluating the success of BEC. Using extensive simulations and six real data examples including scRNA-seq and scATAC-seq datasets with different numbers of batches, batch effect sizes and numbers of cell types, we demonstrate that RBET evaluates the performance of BEC methods more fairly with biologically meaningful insights from data, while other methods may lead to false results. Moreover, RBET is computationally efficient, sensitive to overcorrection and robust to large batch effect sizes. Thus, RBET provides a robust guideline on selecting case-specific BEC method, and the concept of RBET is extendable to other modalities.

SCITUNA: single-cell data integration tool using network alignment

BACKGROUND

As single-cell genomics experiments increase in complexity and scale, the need to integrate multiple datasets has grown. Such integration enhances cellular feature identification by leveraging larger data volumes. However, batch effects-technical variations arising from differences in labs, times, or protocols-pose a significant challenge. Despite numerous proposed batch correction methods, many still have limitations, such as outputting only dimension-reduced data, relying on computationally intensive models, or resulting in overcorrection for batches with diverse cell type composition.

RESULTS

We introduce a novel method for batch effect correction named SCITUNA, a Single-Cell data Integration Tool Using Network Alignment. We perform evaluations on 39 individual batches from four real datasets and a simulated dataset, which include both scRNA-seq and scATAC-seq datasets, spanning multiple organisms and tissues. A thorough comparison of existing batch correction methods using 13 metrics reveals that SCITUNA outperforms current approaches and is successful at preserving biological signals present in the original data. In particular, SCITUNA shows a better performance than the current methods in all the comparisons except for the multiple batch integration of the lung dataset where the difference is 0.004.

CONCLUSION

SCITUNA effectively removes batch effects while retaining the biological signals present in the data. Our extensive experiments reveal that SCITUNA will be a valuable tool for diverse integration tasks.

SPARROW reveals microenvironment-zone-specific cell states in healthy and diseased tissues

Spatially resolved transcriptomics technologies have advanced our understanding of cellular characteristics within tissue contexts. However, current analytical tools often treat cell-type inference and cellular neighborhood identification as separate and hard clustering processes, limiting comparability across scales and samples. SPARROW addresses these challenges by jointly learning latent embeddings and soft clusterings of cell types and cellular organization. It outperformed state-of-the-art methods in cell-type inference and microenvironment zone delineation and uncovered zone-specific cell states in human and mouse tissues that competing methods missed. By integrating spatially resolved transcriptomics and single-cell RNA sequencing (scRNA-seq) data in a shared latent space, SPARROW achieves single-cell spatial resolution and whole-transcriptome coverage, enabling the discovery of both established and unknown microenvironment zone-specific ligand-receptor interactions in the human tonsil. Overall, SPARROW is a computational framework that provides a comprehensive characterization of tissue features across scales, samples, and conditions.

Focal adhesion in the tumour metastasis: from molecular mechanisms to therapeutic targets

The tumour microenvironment is the "hotbed" of tumour cells, providing abundant extracellular support for growth and metastasis. However, the tumour microenvironment is not static and is constantly remodelled by a variety of cellular components, including tumour cells, through mechanical, biological and chemical means to promote metastasis. Focal adhesion plays an important role in cell-extracellular matrix adhesion. An in-depth exploration of the role of focal adhesion in tumour metastasis, especially their contribution at the biomechanical level, is an important direction of current research. In this review, we first summarize the assembly of focal adhesions and explore their kinetics in tumour cells. Then, we describe in detail the role of focal adhesion in various stages of tumour metastasis, especially its key functions in cell migration, invasion, and matrix remodelling. Finally, we describe the anti-tumour strategies targeting focal adhesion and the current progress in the development of some inhibitors against focal adhesion proteins. In this paper, we summarize for the first time that focal adhesion play a positive feedback role in pro-tumour metastatic matrix remodelling by summarizing the five processes of focal adhesion assembly in a multidimensional way. It is beneficial for researchers to have a deeper understanding of the role of focal adhesion in the biological behaviour of tumour metastasis and the potential of focal adhesion as a therapeutic target, providing new ideas for the prevention and treatment of metastases.

Integrative single-cell and spatial transcriptome analysis reveals heterogeneity of human liver progenitor cells

BACKGROUND

Liver progenitor cells (LPCs) with bipotential differentiation capacities are essential for restoring liver homeostasis and hepatocyte population after damage. However, the low proportion and shared markers with epithelial cells make studying LPC heterogeneity difficult, especially in humans. To address this gap, we explored over 259,400 human liver single cells across 4 conditions (fetal, healthy, cirrhotic, and HCC-affected livers).

METHODS

Human liver tissue samples were analyzed using spatial transcriptomics sequencing technologies to describe the heterogeneity of LPCs. Liver tissue was characterized by LPC heterogeneity at single-cell resolution by employing cellular modules, differentiation trajectories, and gene co-expression patterns.

RESULTS

We annotated and identified 1 LPC cluster, 3 LPC subpopulations, and 4 distinct cellular modules, indicating the heterogeneity within LPC and the diversity between LPCs and epithelial cells. LPCs showed spatial colocalization with cholangiocytes and comprised a small proportion (2.95±1.91%) within the merged epithelial cells and LPC populations, exhibiting marked differences in marker expression patterns compared to those in mice. LPCs exhibited distinct cellular states in functional restoration, activation, proliferation, and cell transition. Additionally, the gene co-expression network of LPCs exhibited 3 unique modules, reflecting the distinct connectivity of genes encoding apolipoproteins and heat shock proteins in the gene co-expression network modules.

CONCLUSIONS

Our study provides valuable insights into the multifaceted heterogeneity of human LPCs. Future studies focusing on spatial gene expression dynamics will contribute to our understanding of the spatial arrangement of liver regeneration.

Single-cell analysis of aplastic anemia reveals a convergence of NK and NK-like CD8+ T cells with a disease-associated TCR signature

Immune aplastic anemia (AA) is a life-threatening bone marrow failure disorder driven by an autoimmune T cell attack against hematopoietic stem and progenitor cells (HSPCs). However, the exact autoantigen targets and role of other immune cells in the pathogenesis of AA are unknown. Here, we analyzed a cohort of 218 patients with AA using single-cell RNA and T cell receptor (TCR) αβ sequencing, TCRβ sequencing, flow cytometry, and plasma cytokine profiling. We identified natural killer (NK) cells and CD8+ terminally differentiated effector T (TEMRA) cells expressing NK receptors with AA-associated TCRβ motifs as the most dysregulated immune cell populations in AA bone marrow. Functional coculture experiments using primary HSPCs and immune cells showed that NK cells cannot kill HSPCs alone but may sensitize HSPCs to CD8+ T cell-mediated killing through production of interferons. Furthermore, HSPCs induced activation of T cell clones with CD8+ TEMRA NK-like phenotype in coculture. Our results reveal a convergent phenotype of innate and adaptive immune cells that may drive AA.

Disease diagnostics using machine learning of B cell and T cell receptor sequences

Clinical diagnosis typically incorporates physical examination, patient history, various laboratory tests, and imaging studies but makes limited use of the human immune system's own record of antigen exposures encoded by receptors on B cells and T cells. We analyzed immune receptor datasets from 593 individuals to develop MAchine Learning for Immunological Diagnosis, an interpretive framework to screen for multiple illnesses simultaneously or precisely test for one condition. This approach detects specific infections, autoimmune disorders, vaccine responses, and disease severity differences. Human-interpretable features of the model recapitulate known immune responses to severe acute respiratory syndrome coronavirus 2, influenza, and human immunodeficiency virus, highlight antigen-specific receptors, and reveal distinct characteristics of systemic lupus erythematosus and type-1 diabetes autoreactivity. This analysis framework has broad potential for scientific and clinical interpretation of immune responses.

A single-cell transcriptomic atlas of developing inhibitory neurons reveals expanding and contracting modes of diversification

The cerebral cortex relies on vastly different types of inhibitory neurons to compute. How this diversity emerges during development remains an open question. The rarity of individual inhibitory neuron types often leads to their underrepresentation in single-cell RNA sequencing (scRNAseq) datasets, limiting insights into their developmental trajectories. To address this problem, we developed a computational pipeline to enrich and integrate rare cell types across multiple datasets. Applying this approach to somatostatin-expressing (SST+) inhibitory neurons-the most diverse inhibitory cell class in the cortex-we constructed the Dev-SST-Atlas, a comprehensive resource containing mouse transcriptomic data of over 51,000 SST+ neurons. We identify three principal groups-Martinotti cells (MCs), non-Martinotti cells (nMCs), and long-range projecting neurons (LRPs)-each following distinct diversification trajectories. MCs commit early, with distinct embryonic and neonatal clusters that map directly to adult counterparts. In contrast, nMCs diversify gradually, with each developmental cluster giving rise to multiple adult cell types. LRPs follow a unique 'contracting' mode. Initially, two clusters are present until postnatal day 5 (P5), but by P7, one type is eliminated through programmed cell death, leaving a single surviving population. This transient LRP type is also found in the fetal human cortex, revealing an evolutionarily conserved feature of cortical development. Together, these findings highlight three distinct modes of SST+ neuron diversification-invariant, expanding, and contracting-offering a new framework to understand how the large repertoire of inhibitory neurons emerges during development.

scCobra allows contrastive cell embedding learning with domain adaptation for single cell data integration and harmonization

The rapid advancement of single-cell technologies has created an urgent need for effective methods to integrate and harmonize single-cell data. Technical and biological variations across studies complicate data integration, while conventional tools often struggle with reliance on gene expression distribution assumptions and over-correction. Here, we present scCobra, a deep generative neural network designed to overcome these challenges through contrastive learning with domain adaptation. scCobra effectively mitigates batch effects, minimizes over-correction, and ensures biologically meaningful data integration without assuming specific gene expression distributions. It enables online label transfer across datasets with batch effects, allowing continuous integration of new data without retraining. Additionally, scCobra supports batch effect simulation, advanced multi-omic integration, and scalable processing of large datasets. By integrating and harmonizing datasets from similar studies, scCobra expands the available data for investigating specific biological problems, improving cross-study comparability, and revealing insights that may be obscured in isolated datasets.

scMFG: a single-cell multi-omics integration method based on feature grouping

BACKGROUND

Recent advancements in methodologies and technologies have enabled the simultaneous measurement of multiple omics data, which provides a comprehensive understanding of cellular heterogeneity. However, existing methods have limitations in accurately identifying cell types while maintaining model interpretability, especially in the presence of noise.

METHODS

We propose a novel method called scMFG, which leverages feature grouping and group integration techniques for the integration of single-cell multi-omics data. By organizing features with similar characteristics within each omics layer through feature grouping. Furthermore, scMFG ensures a consistent feature grouping approach across different omics layers, promoting comparability of diverse data types. Additionally, scMFG incorporates a matrix factorization-based approach to enable the integrated results remain interpretable.

RESULTS

We comprehensively evaluated scMFG's performance on four complex real-world datasets generated using diverse sequencing technologies, highlighting its robustness in accurately identifying cell types. Notably, scMFG exhibited superior performance in deciphering cellular heterogeneity at a finer resolution compared to existing methods when applied to simulated datasets. Furthermore, our method proved highly effective in identifying rare cell types, showcasing its robust performance and suitability for detecting low-abundance cellular populations. The interpretability of scMFG was successfully validated through its specific association of outputs with specific cell types or states observed in the neonatal mouse cerebral cortices dataset. Moreover, we demonstrated that scMFG is capable of identifying cell developmental trajectories even in datasets with batch effects.

CONCLUSIONS

Our work presents a robust framework for the analysis of single-cell multi-omics data, advancing our understanding of cellular heterogeneity in a comprehensive and interpretable manner.

EpiFoundation: A Foundation Model for Single-Cell ATAC-seq via Peak-to-Gene Alignment

Foundation models exhibit strong capabilities for downstream tasks by learning generalized representations through self-supervised pre-training on large datasets. While several foundation models have been developed for single-cell RNA-seq (scRNA-seq) data, there is still a lack of models specifically tailored for single-cell ATAC-seq (scATAC-seq), which measures epigenetic information in individual cells. The principal challenge in developing such a model lies in the vast number of scATAC peaks and the significant sparsity of the data, which complicates the formulation of peak-to-peak correlations. To address this challenge, we introduce EpiFoundation, a foundation model for learning cell representations from the high-dimensional and sparse space of peaks. EpiFoundation relies on an innovative cross-modality pre-training procedure with two key technical innovations. First, EpiFoundation exclusively processes the non-zero peak set, thereby enhancing the density of cell-specific information within the input data. Second, EpiFoundation utilizes dense gene expression information to supervise the pre-training process, aligning peak-to-gene correlations. EpiFoundation can handle various types of downstream tasks, including cell-type annotation, batch correction, and gene expression prediction. To train and validate EpiFoundation, we curated MiniAtlas, a dataset of 100,000+ single cells with paired scRNA-seq and scATAC-seq data, along with diverse test sets spanning various tissues and cell types for robust evaluation. EpiFoundation demonstrates state-of-the-art performance across multiple tissues and diverse downstream tasks.

The dynamics of hematopoiesis over the human lifespan

Over a lifetime, hematopoietic stem cells (HSCs) adjust their lineage output to support age-aligned physiology. In model organisms, stereotypic waves of hematopoiesis have been observed corresponding to defined age-biased HSC hallmarks. However, how the properties of hematopoietic stem and progenitor cells change over the human lifespan remains unclear. To address this gap, we profiled individual transcriptome states of human hematopoietic stem and progenitor cells spanning gestation, maturation and aging. Here we define the gene expression networks dictating age-specific differentiation of HSCs and the dynamics of fate decisions and lineage priming throughout life. We additionally identifiy and functionally validate a fetal-specific HSC state with robust engraftment and multilineage capacity. Furthermore, we observe that classification of acute myeloid leukemia against defined transcriptional age states demonstrates that utilization of early life transcriptional programs associates with poor prognosis. Overall, we provide a disease-relevant framework for heterochronic orientation of stem cell ontogeny along the real time axis of the human lifespan.

INSTINCT: Multi-sample integration of spatial chromatin accessibility sequencing data via stochastic domain translation

Recent advances in spatial epigenomic techniques have given rise to spatial assay for transposase-accessible chromatin using sequencing (spATAC-seq) data, enabling the characterization of epigenomic heterogeneity and spatial information simultaneously. Integrative analysis of multiple spATAC-seq samples, for which no method has been developed, allows for effective identification and elimination of unwanted non-biological factors within the data, enabling comprehensive exploration of tissue structures and providing a holistic epigenomic landscape, thereby facilitating the discovery of biological implications and the study of regulatory processes. In this article, we present INSTINCT, a method for multi-sample INtegration of Spatial chromaTIN accessibility sequencing data via stochastiC domain Translation. INSTINCT can efficiently handle the high dimensionality of spATAC-seq data and eliminate the complex noise and batch effects of samples through a stochastic domain translation procedure. We demonstrate the superiority and robustness of INSTINCT in integrating spATAC-seq data across multiple simulated scenarios and real datasets. Additionally, we highlight the advantages of INSTINCT in spatial domain identification, visualization, spot-type annotation, and various downstream analyses, including motif enrichment analysis, expression enrichment analysis, and partitioned heritability analysis.

Benchmark of Data Integration in Single-Cell Proteomics

Single-cell proteomics (SCP) detected based on different technologies always involves batch-specific variations because of differences in sample processing and other potential biases. How to integrate SCP data effectively has become a great challenge. Integration of SCP data not only requires the conservation of true biological variances, but also realizes the removal of unwanted batch effects. In this study, benchmarking analysis of popular data integration methods was conducted to determine the most suitable method for SCP data. To comprehensively evaluate the performance of these integration methods, a novel evaluation system was proposed for integrating SCP data. This evaluation system consists of three objective measures from different perspectives: category (a), the efficacy of correcting batch effects; category (b), the power of conserving biological variances; and category (c), the ability to identify consistent markers. For this comprehensive evaluation, five benchmark data sets under different scenarios (containing substantial proteins, substantial cells, multiple batches, multiple cell types, and unbalanced data) were utilized for selecting the most suitable data integration method. As a result, three methods, ComBat, Scanorama, and Seurat version 3 CCA, were identified as the most recommended methods for integrating SCP data. Overall, this systematic evaluation might provide valuable guidance in choosing the appropriate method for data integration in the SCP.

The Venus score for the assessment of the quality and trustworthiness of biomedical datasets

Biomedical datasets are the mainstays of computational biology and health informatics projects, and can be found on multiple data platforms online or obtained from wet-lab biologists and physicians. The quality and the trustworthiness of these datasets, however, can sometimes be poor, producing bad results in turn, which can harm patients and data subjects. To address this problem, policy-makers, researchers, and consortia have proposed diverse regulations, guidelines, and scores to assess the quality and increase the reliability of datasets. Although generally useful, however, they are often incomplete and impractical. The guidelines of Datasheets for Datasets, in particular, are too numerous; the requirements of the Kaggle Dataset Usability Score focus on non-scientific requisites (for example, including a cover image); and the European Union Artificial Intelligence Act (EU AI Act) sets forth sparse and general data governance requirements, which we tailored to datasets for biomedical AI. Against this backdrop, we introduce our new Venus score to assess the data quality and trustworthiness of biomedical datasets. Our score ranges from 0 to 10 and consists of ten questions that anyone developing a bioinformatics, medical informatics, or cheminformatics dataset should answer before the release. In this study, we first describe the EU AI Act, Datasheets for Datasets, and the Kaggle Dataset Usability Score, presenting their requirements and their drawbacks. To do so, we reverse-engineer the weights of the influential Kaggle Score for the first time and report them in this study. We distill the most important data governance requirements into ten questions tailored to the biomedical domain, comprising the Venus score. We apply the Venus score to twelve datasets from multiple subdomains, including electronic health records, medical imaging, microarray and bulk RNA-seq gene expression, cheminformatics, physiologic electrogram signals, and medical text. Analyzing the results, we surface fine-grained strengths and weaknesses of popular datasets, as well as aggregate trends. Most notably, we find a widespread tendency to gloss over sources of data inaccuracy and noise, which may hinder the reliable exploitation of data and, consequently, research results. Overall, our results confirm the applicability and utility of the Venus score to assess the trustworthiness of biomedical data.

Benchmarking cross-species single-cell RNA-seq data integration methods: towards a cell type tree of life

Cross-species single-cell RNA-seq data hold immense potential for unraveling cell type evolution and transferring knowledge between well-explored and less-studied species. However, challenges arise from interspecific genetic variation, batch effects stemming from experimental discrepancies and inherent individual biological differences. Here, we benchmarked nine data-integration methods across 20 species, encompassing 4.7 million cells, spanning eight phyla and the entire animal taxonomic hierarchy. Our evaluation reveals notable differences between the methods in removing batch effects and preserving biological variance across taxonomic distances. Methods that effectively leverage gene sequence information capture underlying biological variances, while generative model-based approaches excel in batch effect removal. SATURN demonstrates robust performance across diverse taxonomic levels, from cross-genus to cross-phylum, emphasizing its versatility. SAMap excels in integrating species beyond the cross-family level, especially for atlas-level cross-species integration, while scGen shines within or below the cross-class hierarchy. As a result, our analysis offers recommendations and guidelines for selecting suitable integration methods, enhancing cross-species single-cell RNA-seq analyses and advancing algorithm development.

Single-cell Technology in Stem Cell Research

Single-cell technology (SCT), which enables the examination of the fundamental units comprising biological organs, tissues, and cells, has emerged as a powerful tool, particularly in the field of biology, with a profound impact on stem cell research. This innovative technology opens new pathways for acquiring cell-specific data and gaining insights into the molecular pathways governing organ function and biology. SCT is not only frequently used to explore rare and diverse cell types, including stem cells, but it also unveils the intricacies of cellular diversity and dynamics. This perspective, crucial for advancing stem cell research, facilitates non-invasive analyses of molecular dynamics and cellular functions over time. Despite numerous investigations into potential stem cell therapies for genetic disorders, degenerative conditions, and severe injuries, the number of approved stem cell-based treatments remains limited. This limitation is attributed to the various heterogeneities present among stem cell sources, hindering their widespread clinical utilization. Furthermore, stem cell research is intimately connected with cutting-edge technologies, such as microfluidic organoids, CRISPR technology, and cell/tissue engineering. Each strategy developed to overcome the constraints of stem cell research has the potential to significantly impact advanced stem cell therapies. Drawing on the advantages and progress achieved through SCT-based approaches, this study aims to provide an overview of the advancements and concepts associated with the utilization of SCT in stem cell research and its related fields.

Integrating single-cell multimodal epigenomic data using 1D convolutional neural networks

MOTIVATION

Recent experimental developments enable single-cell multimodal epigenomic profiling, which measures multiple histone modifications and chromatin accessibility within the same cell. Such parallel measurements provide exciting new opportunities to investigate how epigenomic modalities vary together across cell types and states. A pivotal step in using these types of data is integrating the epigenomic modalities to learn a unified representation of each cell, but existing approaches are not designed to model the unique nature of this data type. Our key insight is to model single-cell multimodal epigenome data as a multichannel sequential signal.

RESULTS

We developed ConvNet-VAEs, a novel framework that uses one-dimensional (1D) convolutional variational autoencoders (VAEs) for single-cell multimodal epigenomic data integration. We evaluated ConvNet-VAEs on nano-CUT&Tag and single-cell nanobody-tethered transposition followed by sequencing data generated from juvenile mouse brain and human bone marrow. We found that ConvNet-VAEs can perform dimension reduction and batch correction better than previous architectures while using significantly fewer parameters. Furthermore, the performance gap between convolutional and fully connected architectures increases with the number of modalities, and deeper convolutional architectures can increase the performance, while the performance degrades for deeper fully connected architectures. Our results indicate that convolutional autoencoders are a promising method for integrating current and future single-cell multimodal epigenomic datasets.

AVAILABILITY AND IMPLEMENTATION

The source code of VAE models and a demo in Jupyter notebook are available at https://github.com/welch-lab/ConvNetVAE.

Scaling deep identifiable models enables zero-shot characterization of single-cell biological states

How to identify true biological differences across samples while overcoming batch effects has been a persistent challenge in single-cell RNA-seq data analysis, hindering analyses across datasets for transferable biological findings. In this work, we show that scaling up deep identifiable models leads to a surprisingly effective solution for this challenging task. We developed scShift, a deep variational inference framework with theoretical support in disentangling batch-dependent and independent variations. By training the model with compendiums of scRNA-seq atlases, scShift shows remarkable zero-shot capabilities in revealing representations of cell types and biological states in single-cell data while overcoming batch effects. We employed scShift to systematically compare lung fibrosis states across different datasets, tissues and experimental systems. scShift uniquely extrapolates lung fibrosis states to previously unseen post-COVID-19 fibrosis, characterizing universal myeloid-fibrosis signatures, potential repurposing drug targets and fibrosis-associated cell interactions. Evaluations of over 200 trained scShift models demonstrate emergent zero-shot capabilities and a scaling law beyond a transition threshold, with respect to dataset diversity. With its scaling performance on massive single-cell compendiums and exceptional zero-shot capabilities, scShift represents an important advance toward next-generation computational models for single-cell analysis.

Baseline immune state and T-cell clonal kinetics are associated with durable response to CAR-T therapy in large B-cell lymphoma

ABSTRACT

Engineered cellular therapy with CD19-targeting chimeric antigen receptor T cells (CAR-Ts) has revolutionized outcomes for patients with relapsed/refractory large B-cell lymphoma (LBCL), but the cellular and molecular features associated with response remain largely unresolved. We analyzed serial peripheral blood samples ranging from the day of apheresis (day -28/baseline) to 28 days after CAR-T infusion from 50 patients with LBCL treated with axicabtagene ciloleucel by integrating single-cell RNA and T-cell receptor sequencing, flow cytometry, and mass cytometry to characterize features associated with response to CAR-T. Pretreatment patient characteristics associated with response included the presence of B cells and increased absolute lymphocyte count to absolute monocyte count ratio (ALC/AMC). Infusion products from responders were enriched for clonally expanded, highly activated CD8+ T cells. We expanded these observations to 99 patients from the ZUMA-1 cohort and identified a subset of patients with elevated baseline B cells, 80% of whom were complete responders. We integrated B-cell proportion ≥0.5% and ALC/AMC ≥1.2 into a 2-factor predictive model and applied this model to the ZUMA-1 cohort. Estimated progression-free survival at 1 year in patients meeting 1 or both criteria was 65% vs 31% for patients meeting neither criterion. Our results suggest that patients' immunologic state at baseline affects the likelihood of response to CAR-T through both modulation of the T-cell apheresis product composition and promoting a more favorable circulating immune compartment before therapy. These baseline immunologic features, measured readily in the clinical setting before CAR-T, can be applied to predict response to therapy.

Computational analysis of the functional impact of MHC-II-expressing triple-negative breast cancer

The tumor microenvironment (TME) plays a crucial role in tumor progression and immunoregulation. Major histocompatibility complex class II (MHC-II) is essential for immune surveillance within the TME. While MHC-II genes are typically expressed by professional antigen-presenting cells, they are also expressed in tumor cells, potentially facilitating antitumor immune responses. To understand the role of MHC-II-expressing tumor cells, we analyzed triple-negative breast cancer (TNBC), an aggressive subtype with poor prognosis and limited treatment options, using public bulk RNA-seq, single-cell RNA-seq, and spatial transcriptomics datasets. Our analysis revealed a distinct tumor subpopulation that upregulates MHC-II genes and actively interacts with immune cells. We implicated that this subpopulation is preferentially present in proximity to regions in immune infiltration of TNBC patient cohorts with a better prognosis, suggesting the functional importance of MHC-II-expressing tumor cells in modulating the immune landscape and influencing patient survival outcomes. Remarkably, we identified a prognostic signature comprising 40 significant genes in the MHC-II-expressing tumors in which machine leaning models with the signature successfully predicted patient survival outcomes and the degree of immune infiltration. This study advances our understanding of the immunological basis of cancer progression and suggests promising new directions for therapeutic strategies.

Recovery of biological signals lost in single-cell batch integration with CellANOVA

Data integration to align cells across batches has become a cornerstone of single-cell data analysis, critically affecting downstream results. Currently, there are no guidelines for when the biological differences between samples are separable from batch effects. Here we show that current paradigms for single-cell data integration remove biologically meaningful variation and introduce distortion. We present a statistical model and computationally scalable algorithm, CellANOVA (cell state space analysis of variance), that harnesses experimental design to explicitly recover biological signals that are erased during single-cell data integration. CellANOVA uses a 'pool-of-controls' design concept, applicable across diverse settings, to separate unwanted variation from biological variation of interest and allow the recovery of subtle biological signals. We apply CellANOVA to diverse contexts and validate the recovered biological signals by orthogonal assays. In particular, we show that CellANOVA is effective in the challenging case of single-cell and single-nucleus data integration, where it recovers subtle biological signals that can be validated and replicated by external data.

Multiomics reveals age-dependent metabolic reprogramming of macrophages by wound bed niche secreted signals

The cellular metabolism of macrophages depends on tissue niches and can control macrophage inflammatory or resolving phenotypes. Yet, the identity of signals within tissue niches that control macrophage metabolism is not well understood. Here, using single-cell RNA sequencing of macrophages in early mouse wounds, we find that, rather than gene expression of canonical inflammatory or resolving polarization markers, metabolic gene expression defines distinct populations of early wound macrophages. Single-cell secretomics and transcriptomics identify inflammatory and resolving cytokines expressed by early wound macrophages, and we show that these signals drive metabolic inputs and mitochondrial metabolism in an age-dependent manner. We show that aging alters the metabolome of early wound macrophages and rewires their metabolism from mitochondria to glycolysis. We further show that macrophage-derived Chi3l3 and IGF-1 can induce metabolic inputs and mitochondrial mass/metabolism in aged and bone marrow-derived macrophages. Together, these findings reveal that macrophage-derived signals drive the mitochondrial metabolism of macrophages within early wounds in an age-dependent manner and have implications for inflammatory diseases, chronic injuries, and age-related inflammatory diseases.

In Brief

This study reveals that macrophage subsets in early inflammatory stages of skin wound healing are defined by their metabolic profiles rather than polarization phenotype. Using single-cell secretomics, we establish key macrophage cytokines that comprise the in vivo wound niche and drive mitochondrial-based metabolism. Aging significantly alters macrophage heterogeneity and increases glycolytic metabolism, which can be restored to OxPHOS-based metabolism with young niche cytokines. These findings highlight the importance of the tissue niche in driving macrophage phenotypes, with implications for aging-related impairments in wound healing.

Highlights

Single cell transcriptional analysis reveals that reveals that metabolic gene expression identifies distinct macrophage populations in early skin wounds.Single-cell secretomic data show that young macrophages contribute to the wound bed niche by secreting molecules such as IGF-1 and Chi3l3.Old wound macrophages display altered metabolomics, elevated glycolytic metabolism and glucose uptake, and reduced lipid uptake and mitochondrial mass/metabolism.Chi3l3 but not IGF-1 secretion is altered in macrophages in an age dependent manner.Chi3l3 can restore mitochondrial mass/metabolism in aged macrophages.

Benchmarking algorithms for single-cell multi-omics prediction and integration

The development of single-cell multi-omics technology has greatly enhanced our understanding of biology, and in parallel, numerous algorithms have been proposed to predict the protein abundance and/or chromatin accessibility of cells from single-cell transcriptomic information and to integrate various types of single-cell multi-omics data. However, few studies have systematically compared and evaluated the performance of these algorithms. Here, we present a benchmark study of 14 protein abundance/chromatin accessibility prediction algorithms and 18 single-cell multi-omics integration algorithms using 47 single-cell multi-omics datasets. Our benchmark study showed overall totalVI and scArches outperformed the other algorithms for predicting protein abundance, and LS_Lab was the top-performing algorithm for the prediction of chromatin accessibility in most cases. Seurat, MOJITOO and scAI emerge as leading algorithms for vertical integration, whereas totalVI and UINMF excel beyond their counterparts in both horizontal and mosaic integration scenarios. Additionally, we provide a pipeline to assist researchers in selecting the optimal multi-omics prediction and integration algorithm.

Understanding the gastrointestinal microbiome in systemic sclerosis: methodological advancements and emerging research

PURPOSE OF REVIEW

This review highlights the role of the gastrointestinal (GI) microbiome in systemic sclerosis (SSc). We describe techniques for evaluating the GI microbiome in humans, and emerging research linking GI microbiome alterations (i.e., dysbiosis) and distinct SSc clinical manifestations. We also address the evolving treatment landscape targeting dysbiosis in SSc.

RECENT FINDINGS

Recent literature brings into focus the complex relationship between the GI microbiome and SSc pathogenesis. Advanced techniques (e.g., shotgun metagenomics, meta-transcriptomics) provide deeper insights into microbial taxonomy and active gene expression, exposing dysbiosis as a potential driver of SSc. New studies demonstrate that SSc patients who possess specific SSc clinical features, (e.g., interstitial lung disease), have unique GI microbiome profiles.

SUMMARY

Dysbiosis is associated with specific clinical features in patients with SSc. New tools for studying the GI microbiome have furthered our understanding of the relationship between dysbiosis and SSc complications. Therapeutic avenues such as dietary adjustments, probiotics, antibiotics, mindfulness practices, and fecal transplants offer potential for managing SSc and preventing its progression through GI microbiome modulation. By clarifying what is known about the relationship between the GI dysbiosis, GI dysfunction, and SSc, this review enhances our understanding of SSc pathogenesis and proposes targeted interventions.

IRescue: uncertainty-aware quantification of transposable elements expression at single cell level

Transposable elements (TEs) are mobile DNA repeats known to shape the evolution of eukaryotic genomes. In complex organisms, they exhibit tissue-specific transcription. However, understanding their role in cellular diversity across most tissues remains a challenge, when employing single-cell RNA sequencing (scRNA-seq), due to their widespread presence and genetic similarity. To address this, we present IRescue (Interspersed Repeats single-cell quantifier), a software capable of estimating the expression of TE subfamilies at the single-cell level. IRescue incorporates a unique UMI deduplication algorithm to rectify sequencing errors and employs an Expectation-Maximization procedure to effectively redistribute the counts of multi-mapping reads. Our study showcases the precision of IRescue through analysis of both simulated and real single cell and nuclei RNA-seq data from human colorectal cancer, brain, skin aging, and PBMCs during SARS-CoV-2 infection and recovery. By linking the expression patterns of TE signatures to specific conditions and biological contexts, we unveil insights into their potential roles in cellular heterogeneity and disease progression.

Assessing and mitigating batch effects in large-scale omics studies

Batch effects in omics data are notoriously common technical variations unrelated to study objectives, and may result in misleading outcomes if uncorrected, or hinder biomedical discovery if over-corrected. Assessing and mitigating batch effects is crucial for ensuring the reliability and reproducibility of omics data and minimizing the impact of technical variations on biological interpretation. In this review, we highlight the profound negative impact of batch effects and the urgent need to address this challenging problem in large-scale omics studies. We summarize potential sources of batch effects, current progress in evaluating and correcting them, and consortium efforts aiming to tackle them.

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 .

Trajectory-centric framework TrajAtlas reveals multi-scale differentiation heterogeneity among cells, genes, and gene modules in osteogenesis

Osteoblasts, the key cells responsible for bone formation and the maintenance of skeletal integrity, originate from a diverse array of progenitor cells. However, the mechanisms underlying osteoblast differentiation from these multiple osteoprogenitors remain poorly understood. To address this knowledge gap, we developed a comprehensive framework to investigate osteoblast differentiation at multiple scales, encompassing cells, genes, and gene modules. We constructed a reference atlas focused on differentiation, which incorporates various osteoprogenitors and provides a seven-level cellular taxonomy. To reconstruct the differentiation process, we developed a model that identifies the transcription factors and pathways involved in differentiation from different osteoprogenitors. Acknowledging that covariates such as age and tissue type can influence differentiation, we created an algorithm to detect differentially expressed genes throughout the differentiation process. Additionally, we implemented methods to identify conserved pseudotemporal gene modules across multiple samples. Overall, our framework systematically addresses the heterogeneity observed during osteoblast differentiation from diverse sources, offering novel insights into the complexities of bone formation and serving as a valuable resource for understanding osteogenesis.

Single-cell sequencing to multi-omics: technologies and applications

Cells, as the fundamental units of life, contain multidimensional spatiotemporal information. Single-cell RNA sequencing (scRNA-seq) is revolutionizing biomedical science by analyzing cellular state and intercellular heterogeneity. Undoubtedly, single-cell transcriptomics has emerged as one of the most vibrant research fields today. With the optimization and innovation of single-cell sequencing technologies, the intricate multidimensional details concealed within cells are gradually unveiled. The combination of scRNA-seq and other multi-omics is at the forefront of the single-cell field. This involves simultaneously measuring various omics data within individual cells, expanding our understanding across a broader spectrum of dimensions. Single-cell multi-omics precisely captures the multidimensional aspects of single-cell transcriptomes, immune repertoire, spatial information, temporal information, epitopes, and other omics in diverse spatiotemporal contexts. In addition to depicting the cell atlas of normal or diseased tissues, it also provides a cornerstone for studying cell differentiation and development patterns, disease heterogeneity, drug resistance mechanisms, and treatment strategies. Herein, we review traditional single-cell sequencing technologies and outline the latest advancements in single-cell multi-omics. We summarize the current status and challenges of applying single-cell multi-omics technologies to biological research and clinical applications. Finally, we discuss the limitations and challenges of single-cell multi-omics and potential strategies to address them.

Thinking points for effective batch correction on biomedical data

Batch effects introduce significant variability into high-dimensional data, complicating accurate analysis and leading to potentially misleading conclusions if not adequately addressed. Despite technological and algorithmic advancements in biomedical research, effectively managing batch effects remains a complex challenge requiring comprehensive considerations. This paper underscores the necessity of a flexible and holistic approach for selecting batch effect correction algorithms (BECAs), advocating for proper BECA evaluations and consideration of artificial intelligence-based strategies. We also discuss key challenges in batch effect correction, including the importance of uncovering hidden batch factors and understanding the impact of design imbalance, missing values, and aggressive correction. Our aim is to provide researchers with a robust framework for effective batch effects management and enhancing the reliability of high-dimensional data analyses.

Batch-effect correction in single-cell RNA sequencing data using JIVE

Motivation

In single-cell RNA sequencing analysis, addressing batch effects-technical artifacts stemming from factors such as varying sequencing technologies, equipment, and capture times-is crucial. These factors can cause unwanted variation and obfuscate the underlying biological signal of interest. The joint and individual variation explained (JIVE) method can be used to extract shared biological patterns from multi-source sequencing data while adjusting for individual non-biological variations (i.e. batch effect). However, its current implementation is originally designed for bulk sequencing data, making it computationally infeasible for large-scale single-cell sequencing datasets.

Results

In this study, we enhance JIVE for large-scale single-cell data by boosting its computational efficiency. Additionally, we introduce a novel application of JIVE for batch-effect correction on multiple single-cell sequencing datasets. Our enhanced method aims to decompose single-cell sequencing datasets into a joint structure capturing the true biological variability and individual structures, which capture technical variability within each batch. This joint structure is then suitable for use in downstream analyses. We benchmarked the results against four popular tools, Seurat v5, Harmony, LIGER, and Combat-seq, which were developed for this purpose. JIVE performed best in terms of preserving cell-type effects and in scenarios in which the batch sizes are balanced.

Availability and implementation

The JIVE implementation used for this analysis can be found at https://github.com/oconnell-statistics-lab/scJIVE.

New frameworks for hematopoiesis derived from single-cell genomics

ABSTRACT

Recent advancements in single-cell genomics have enriched our understanding of hematopoiesis, providing intricate details about hematopoietic stem cell biology, differentiation, and lineage commitment. Technological advancements have highlighted extensive heterogeneity of cell populations and continuity of differentiation routes. Nevertheless, intermediate "attractor" states signify structure in stem and progenitor populations that link state transition dynamics to fate potential. We discuss how innovative model systems quantify lineage bias and how stress accelerates differentiation, thereby reducing fate plasticity compared with native hematopoiesis. We conclude by offering our perspective on the current model of hematopoiesis and discuss how a more precise understanding can translate to strategies that extend healthy hematopoiesis and prevent disease.

Human breastmilk memory T cells throughout lactation manifest activated tissue-oriented profile with prominent regulation

Breastfeeding provides important immunological benefits to the neonate, but how the different immunoactive components in breastmilk contribute to immunity remains poorly understood. Here, we characterized human breastmilk T cells using single-cell RNA-Seq and flow cytometry. Breastmilk contained predominantly memory T cells, with expression of immune signaling genes, high proliferation, and an effector Th1/cytotoxic profile with high cytokine production capacities. Elevated activation was balanced by an enriched Treg population and immune regulatory markers in conventional memory T cells. Gene and surface expression of tissue-residency markers indicate that breastmilk T cells represented tissue-adapted rather than circulatory T cells. In addition, breastmilk T cells had a broad homing profile and higher activation markers in these migratory subsets. The partly overlapping transcriptome profile between breastmilk and breast tissue T cells, particularly cytotoxic T cells, might support a role in local immune defense in the mammary gland. However, unique features of breastmilk, such as Tregs, might imply an additional role in neonatal immune support. We found some correlations between the breastmilk T cell profile and clinical parameters, most notably with maternal and household factors. Together, our data suggest that breastmilk contains an adapted T cell population that exerts their function in specific tissue sites.

NMR metabolomics-guided DNA methylation mortality predictors

BACKGROUND

1H-NMR metabolomics and DNA methylation in blood are widely known biomarkers predicting age-related physiological decline and mortality yet exert mutually independent mortality and frailty signals.

METHODS

Leveraging multi-omics data in four Dutch population studies (N = 5238, ∼40% of which male) we investigated whether the mortality signal captured by 1H-NMR metabolomics could guide the construction of DNA methylation-based mortality predictors.

FINDINGS

We trained DNA methylation-based surrogates for 64 metabolomic analytes and found that analytes marking inflammation, fluid balance, or HDL/VLDL metabolism could be accurately reconstructed using DNA-methylation assays. Interestingly, a previously reported multi-analyte score indicating mortality risk (MetaboHealth) could also be accurately reconstructed. Sixteen of our derived surrogates, including the MetaboHealth surrogate, showed significant associations with mortality, independent of relevant covariates.

INTERPRETATION

The addition of our metabolic analyte-derived surrogates to the well-established epigenetic clock GrimAge demonstrates that our surrogates potentially represent valuable mortality signal.

FUNDING

BBMRI-NL, X-omics, VOILA, Medical Delta, NWO, ERC.

A brain cell atlas integrating single-cell transcriptomes across human brain regions

While single-cell technologies have greatly advanced our comprehension of human brain cell types and functions, studies including large numbers of donors and multiple brain regions are needed to extend our understanding of brain cell heterogeneity. Integrating atlas-level single-cell data presents a chance to reveal rare cell types and cellular heterogeneity across brain regions. Here we present the Brain Cell Atlas, a comprehensive reference atlas of brain cells, by assembling single-cell data from 70 human and 103 mouse studies of the brain throughout major developmental stages across brain regions, covering over 26.3 million cells or nuclei from both healthy and diseased tissues. Using machine-learning based algorithms, the Brain Cell Atlas provides a consensus cell type annotation, and it showcases the identification of putative neural progenitor cells and a cell subpopulation of PCDH9high microglia in the human brain. We demonstrate the gene regulatory difference of PCDH9high microglia between hippocampus and prefrontal cortex and elucidate the cell-cell communication network. The Brain Cell Atlas presents an atlas-level integrative resource for comparing brain cells in different environments and conditions within the Human Cell Atlas.

Graph Fourier transform for spatial omics representation and analyses of complex organs

Spatial omics technologies decipher functional components of complex organs at cellular and subcellular resolutions. We introduce Spatial Graph Fourier Transform (SpaGFT) and apply graph signal processing to a wide range of spatial omics profiling platforms to generate their interpretable representations. This representation supports spatially variable gene identification and improves gene expression imputation, outperforming existing tools in analyzing human and mouse spatial transcriptomics data. SpaGFT can identify immunological regions for B cell maturation in human lymph nodes Visium data and characterize variations in secondary follicles using in-house human tonsil CODEX data. Furthermore, it can be integrated seamlessly into other machine learning frameworks, enhancing accuracy in spatial domain identification, cell type annotation, and subcellular feature inference by up to 40%. Notably, SpaGFT detects rare subcellular organelles, such as Cajal bodies and Set1/COMPASS complexes, in high-resolution spatial proteomics data. This approach provides an explainable graph representation method for exploring tissue biology and function.

An end-to-end deep learning method for mass spectrometry data analysis to reveal disease-specific metabolic profiles

Untargeted metabolomic analysis using mass spectrometry provides comprehensive metabolic profiling, but its medical application faces challenges of complex data processing, high inter-batch variability, and unidentified metabolites. Here, we present DeepMSProfiler, an explainable deep-learning-based method, enabling end-to-end analysis on raw metabolic signals with output of high accuracy and reliability. Using cross-hospital 859 human serum samples from lung adenocarcinoma, benign lung nodules, and healthy individuals, DeepMSProfiler successfully differentiates the metabolomic profiles of different groups (AUC 0.99) and detects early-stage lung adenocarcinoma (accuracy 0.961). Model flow and ablation experiments demonstrate that DeepMSProfiler overcomes inter-hospital variability and effects of unknown metabolites signals. Our ensemble strategy removes background-category phenomena in multi-classification deep-learning models, and the novel interpretability enables direct access to disease-related metabolite-protein networks. Further applying to lipid metabolomic data unveils correlations of important metabolites and proteins. Overall, DeepMSProfiler offers a straightforward and reliable method for disease diagnosis and mechanism discovery, enhancing its broad applicability.

Cell-vision fusion: A Swin transformer-based approach for predicting kinase inhibitor mechanism of action from Cell Painting data

Image-based profiling of the cellular response to drug compounds has proven effective at characterizing the morphological changes resulting from perturbation experiments. As data availability increases, however, there are growing demands for novel deep-learning methods. We applied the SwinV2 computer vision architecture to predict the mechanism of action of 10 kinase inhibitor compounds directly from Cell Painting images. This method outperforms the standard approach of using image-based profiles (IBP)-multidimensional feature set representations generated by bioimaging software. Furthermore, our fusion approach-cell-vision fusion, combining three different data modalities, images, IBPs, and chemical structures-achieved 69.79% accuracy and 70.56% F1 score, 4.20% and 5.49% higher, respectively, than the best-performing IBP method. We provide three techniques, specific to Cell Painting images, which enable deep-learning architectures to train effectively and demonstrate approaches to combat the significant batch effects present in large Cell Painting datasets.

A Cell Cycle-Aware Network for Data Integration and Label Transferring of Single-Cell RNA-Seq and ATAC-Seq

In recent years, the integration of single-cell multi-omics data has provided a more comprehensive understanding of cell functions and internal regulatory mechanisms from a non-single omics perspective, but it still suffers many challenges, such as omics-variance, sparsity, cell heterogeneity, and confounding factors. As it is known, the cell cycle is regarded as a confounder when analyzing other factors in single-cell RNA-seq data, but it is not clear how it will work on the integrated single-cell multi-omics data. Here, a cell cycle-aware network (CCAN) is developed to remove cell cycle effects from the integrated single-cell multi-omics data while keeping the cell type-specific variations. This is the first computational model to study the cell-cycle effects in the integration of single-cell multi-omics data. Validations on several benchmark datasets show the outstanding performance of CCAN in a variety of downstream analyses and applications, including removing cell cycle effects and batch effects of scRNA-seq datasets from different protocols, integrating paired and unpaired scRNA-seq and scATAC-seq data, accurately transferring cell type labels from scRNA-seq to scATAC-seq data, and characterizing the differentiation process from hematopoietic stem cells to different lineages in the integration of differentiation data.

Immunological characterization and comparison of children with COVID-19 from their adult counterparts at single-cell resolution

Introduction

The immunological characteristics that could protect children with coronavirus disease 2019 (COVID-19) from severe or fatal illnesses have not been fully understood yet.

Methods

Here, we performed single-cell RNA sequencing (scRNA-seq) analysis on peripheral blood samples of 15 children (8 with COVID-19) and compared them to 18 adults (13 with COVID-19).

Results

The child-adult integrated single cell data indicated that children with the disease presented a restrained response to type I interferon in most of the major immune cell types, along with suppression of upstream interferon regulatory factor and toll-like receptor expression in monocytes, which was confirmed by in vitro interferon stimulation assays. Unlike adult patients, children with COVID-19 showed lower frequencies of activated proinflammatory CD14+ monocytes, possibly explaining the rareness of cytokine storm in them. Notably, natural killer (NK) cells in pediatric patients displayed potent cytotoxicity with a rich expression of cytotoxic molecules and upregulated cytotoxic pathways, whereas the cellular senescence, along with the Notch signaling pathway, was significantly downregulated in NK cells, all suggesting more robust cytotoxicity in NK cells of children than adult patients that was further confirmed by CD107a degranulation assays. Lastly, a modest adaptive immune response was evident with more naïve T cells but less activated and proliferated T cells while less naïve B cells but more activated B cells in children over adult patients.

Conclusion

Conclusively, this preliminary study revealed distinct cell frequency and activation status of major immune cell types, particularly more robust NK cell cytotoxicity in PBMC that might help protect children from severe COVID-19.

DeepComBat: A statistically motivated, hyperparameter-robust, deep learning approach to harmonization of neuroimaging data

Neuroimaging data acquired using multiple scanners or protocols are increasingly available. However, such data exhibit technical artifacts across batches which introduce confounding and decrease reproducibility. This is especially true when multi-batch data are analyzed using complex downstream models which are more likely to pick up on and implicitly incorporate batch-related information. Previously proposed image harmonization methods have sought to remove these batch effects; however, batch effects remain detectable in the data after applying these methods. We present DeepComBat, a deep learning harmonization method based on a conditional variational autoencoder and the ComBat method. DeepComBat combines the strengths of statistical and deep learning methods in order to account for the multivariate relationships between features while simultaneously relaxing strong assumptions made by previous deep learning harmonization methods. As a result, DeepComBat can perform multivariate harmonization while preserving data structure and avoiding the introduction of synthetic artifacts. We apply this method to cortical thickness measurements from a cognitive-aging cohort and show DeepComBat qualitatively and quantitatively outperforms existing methods in removing batch effects while preserving biological heterogeneity. Additionally, DeepComBat provides a new perspective for statistically motivated deep learning harmonization methods.