SPACEL: deep learning-based characterization of spatial transcriptome architectures

Cell-type deconvolution methods for spatial transcriptomics

Spatial transcriptomics is a powerful method for studying the spatial organization of cells, which is a critical feature in the development, function and evolution of multicellular life. However, sequencing-based spatial transcriptomics has not yet achieved cellular-level resolution, so advanced deconvolution methods are needed to infer cell-type contributions at each location in the data. Recent progress has led to diverse tools for cell-type deconvolution that are helping to describe tissue architectures in health and disease. In this Review, we describe the varied types of cell-type deconvolution methods for spatial transcriptomics, contrast their capabilities and summarize them in a web-based, interactive table to enable more efficient method selection.

Unravelling tumour spatiotemporal heterogeneity using spatial multimodal data

Analysing the genome, epigenome, transcriptome, proteome, and metabolome within the spatial context of cells has transformed our understanding of tumour spatiotemporal heterogeneity. Advances in spatial multi-omics technologies now reveal complex molecular interactions shaping cellular behaviour and tissue dynamics. This review highlights key technologies and computational methods that have advanced spatial domain identification and their pseudo-relations, as well as inference of intra- and inter-cellular molecular networks that drive disease progression. We also discuss strategies to address major challenges, including data sparsity, high-dimensionality, scalability, and heterogeneity. Furthermore, we outline how spatial multi-omics enables novel insights into disease mechanisms, advancing precision medicine and informing targeted therapies. KEY POINTS: Advancements in spatial multi-omics facilitate our understanding of tumour spatiotemporal heterogeneity. AI-driven multimodal models uncover complex molecular interactions that underlie cellular behaviours and tissue dynamics. Combining multi-omics technologies and AI-enabled bioinformatics tools helps predict critical disease stages, such as pre-cancer, advancing precision medicine, and informing targeted therapeutic strategies.

KanCell: dissecting cellular heterogeneity in biological tissues through integrated single-cell and spatial transcriptomics

KanCell is a deep learning model based on Kolmogorov-Arnold networks (KAN) designed to enhance cellular heterogeneity analysis by integrating single-cell RNA sequencing and spatial transcriptomics (ST) data. ST technologies provide insights into gene expression within tissue context, revealing cellular interactions and microenvironments. To fully leverage this potential, effective computational models are crucial. We evaluate KanCell on both simulated and real datasets from technologies such as STARmap, Slide-seq, Visium, and Spatial Transcriptomics. Our results demonstrate that KanCell outperforms existing methods across metrics like PCC, SSIM, COSSIM, RMSE, JSD, ARS, and ROC, with robust performance under varying cell numbers and background noise. Real-world applications on human lymph nodes, hearts, melanoma, breast cancer, dorsolateral prefrontal cortex, and mouse embryo brains confirmed its reliability. Compared with traditional approaches, KanCell effectively captures non-linear relationships and optimizes computational efficiency through KAN, providing an accurate and efficient tool for ST. By improving data accuracy and resolving cell type composition, KanCell reveals cellular heterogeneity, clarifies disease microenvironments, and identifies therapeutic targets, addressing complex biological challenges.

Scalable image-based visualization and alignment of spatial transcriptomics datasets

We present the "spatial transcriptomics imaging framework" (STIM), an imaging-based computational framework focused on visualizing and aligning high-throughput spatial sequencing datasets. STIM is built on the powerful, scalable ImgLib2 and BigDataViewer (BDV) image data frameworks and thus enables novel development or transfer of existing computer vision techniques to the sequencing domain characterized by datasets with irregular measurement-spacing and arbitrary spatial resolution, such as spatial transcriptomics data generated by multiplexed targeted hybridization or spatial sequencing technologies. We illustrate STIM's capabilities by representing, interactively visualizing, 3D rendering, automatically registering, and segmenting publicly available spatial sequencing data from 13 serial sections of mouse brain tissue and from 19 sections of a human metastatic lymph node. We demonstrate that the simplest alignment mode of STIM achieves human-level accuracy.

Artificial intelligence in traditional Chinese medicine: advances in multi-metabolite multi-target interaction modeling

Traditional Chinese Medicine (TCM) utilizes multi-metabolite and multi-target interventions to address complex diseases, providing advantages over single-target therapies. However, the active metabolites, therapeutic targets, and especially the combination mechanisms remain unclear. The integration of advanced data analysis and nonlinear modeling capabilities of artificial intelligence (AI) is driving the transformation of TCM into precision medicine. This review concentrates on the application of AI in TCM target prediction, including multi-omics techniques, TCM-specialized databases, machine learning (ML), deep learning (DL), and cross-modal fusion strategies. It also critically analyzes persistent challenges such as data heterogeneity, limited model interpretability, causal confounding, and insufficient robustness validation in practical applications. To enhance the reliability and scalability of AI in TCM target prediction, future research should prioritize continuous optimization of the AI algorithms using zero-shot learning, end-to-end architectures, and self-supervised contrastive learning.

SpaMask: Dual masking graph autoencoder with contrastive learning for spatial transcriptomics

Understanding the spatial locations of cell within tissues is crucial for unraveling the organization of cellular diversity. Recent advancements in spatial resolved transcriptomics (SRT) have enabled the analysis of gene expression while preserving the spatial context within tissues. Spatial domain characterization is a critical first step in SRT data analysis, providing the foundation for subsequent analyses and insights into biological implications. Graph neural networks (GNNs) have emerged as a common tool for addressing this challenge due to the structural nature of SRT data. However, current graph-based deep learning approaches often overlook the instability caused by the high sparsity of SRT data. Masking mechanisms, as an effective self-supervised learning strategy, can enhance the robustness of these models. To this end, we propose SpaMask, dual masking graph autoencoder with contrastive learning for SRT analysis. Unlike previous GNNs, SpaMask masks a portion of spot nodes and spot-to-spot edges to enhance its performance and robustness. SpaMask combines Masked Graph Autoencoders (MGAE) and Masked Graph Contrastive Learning (MGCL) modules, with MGAE using node masking to leverage spatial neighbors for improved clustering accuracy, while MGCL applies edge masking to create a contrastive loss framework that tightens embeddings of adjacent nodes based on spatial proximity and feature similarity. We conducted a comprehensive evaluation of SpaMask on eight datasets from five different platforms. Compared to existing methods, SpaMask achieves superior clustering accuracy and effective batch correction.

Optimizing Xenium In Situ data utility by quality assessment and best-practice analysis workflows

The Xenium In Situ platform is a new spatial transcriptomics product commercialized by 10x Genomics, capable of mapping hundreds of genes in situ at subcellular resolution. Given the multitude of commercially available spatial transcriptomics technologies, recommendations in choice of platform and analysis guidelines are increasingly important. Herein, we explore 25 Xenium datasets generated from multiple tissues and species, comparing scalability, resolution, data quality, capacities and limitations with eight other spatially resolved transcriptomics technologies and commercial platforms. In addition, we benchmark the performance of multiple open-source computational tools, when applied to Xenium datasets, in tasks including preprocessing, cell segmentation, selection of spatially variable features and domain identification. This study serves as an independent analysis of the performance of Xenium, and provides best practices and recommendations for analysis of such datasets.

Chinese Medicine in the Era of Artificial Intelligence: Challenges and Development Prospects

Traditional Chinese medicine (TCM) has protected the health of Chinese people for thousands of years. With the rapid development of artificial intelligence (AI), various fields of TCM are facing both opportunities and challenges. This review discusses the development prospects and challenges of Chinese medicine in the AI era, emphasizing that AI, as an important tool in the process of Chinese medicine healthcare services, can assist doctors in making objective, rational and professional treatment decisions, and that AI has a strong potential for development in the field of Chinese medicine. However, the emotions, complex thoughts, and humanistic values of doctors are qualities that AI is currently unable to realize, so as the dominant player, the doctor is indispensable to the medical process. By summarizing and analyzing the current development status of AI in diagnosis, drug research, health management and education in TCM, this paper reveals the development prospects and potential risks of combining TCM with AI, and suggests that AI is an important aid for modernizing and improving the quality of TCM medical care in a coordinated manner.

Thor: a platform for cell-level investigation of spatial transcriptomics and histology

Spatial transcriptomics integrates transcriptomics data with histological tissue images, offering deeper insights into cellular organization and molecular functions. However, existing computational platforms mainly focus on genomic analysis, leaving a gap in the seamless integration of genomic and image analysis. To address this, we introduce Thor, a comprehensive computational platform for multi-modal analysis of spatial transcriptomics and histological images. Thor utilizes an anti-shrinking Markov diffusion method to infer single-cell spatial transcriptomes from spot-level data, effectively integrating cell morphology with spatial transcriptomics. The platform features 10 modules designed for cell-level genomic and image analysis. Additionally, we present Mjolnir, a web-based tool for interactive tissue analysis using vivid gigapixel images that display information on histology, gene expression, pathway enrichment, and immune response. Thor's accuracy was validated through simulations and ISH, MERFISH, Xenium, and Stereo-seq datasets. To demonstrate its versatility, we applied Thor for joint genomic-histology analysis across various datasets. In in-house heart failure patient samples, Thor identified a regenerative signature in heart failure, with protein presence confirmed in blood vessels through immunofluorescence staining. Thor also revealed the layered structure of the mouse olfactory bulb, performed unbiased screening of breast cancer hallmarks, elucidated the heterogeneity of immune responses, and annotated fibrotic regions in multiple heart failure zones using a semi-supervised approach. Furthermore, Thor imputed high-resolution spatial transcriptomics data in an in-house bladder cancer sample sequenced using Visium HD, uncovering stronger spatial patterns that align more closely with histology. Bridging the gap between genomic and image analysis in spatial biology, Thor offers a powerful tool for comprehensive cellular and molecular analysis.

Graph neural networks for single-cell omics data: a review of approaches and applications

Rapid advancement of sequencing technologies now allows for the utilization of precise signals at single-cell resolution in various omics studies. However, the massive volume, ultra-high dimensionality, and high sparsity nature of single-cell data have introduced substantial difficulties to traditional computational methods. The intricate non-Euclidean networks of intracellular and intercellular signaling molecules within single-cell datasets, coupled with the complex, multimodal structures arising from multi-omics joint analysis, pose significant challenges to conventional deep learning operations reliant on Euclidean geometries. Graph neural networks (GNNs) have extended deep learning to non-Euclidean data, allowing cells and their features in single-cell datasets to be modeled as nodes within a graph structure. GNNs have been successfully applied across a broad range of tasks in single-cell data analysis. In this survey, we systematically review 107 successful applications of GNNs and their six variants in various single-cell omics tasks. We begin by outlining the fundamental principles of GNNs and their six variants, followed by a systematic review of GNN-based models applied in single-cell epigenomics, transcriptomics, spatial transcriptomics, proteomics, and multi-omics. In each section dedicated to a specific omics type, we have summarized the publicly available single-cell datasets commonly utilized in the articles reviewed in that section, totaling 77 datasets. Finally, we summarize the potential shortcomings of current research and explore directions for future studies. We anticipate that this review will serve as a guiding resource for researchers to deepen the application of GNNs in single-cell omics.

From morphology to single-cell molecules: high-resolution 3D histology in biomedicine

High-resolution three-dimensional (3D) tissue analysis has emerged as a transformative innovation in the life sciences, providing detailed insights into the spatial organization and molecular composition of biological tissues. This review begins by tracing the historical milestones that have shaped the development of high-resolution 3D histology, highlighting key breakthroughs that have facilitated the advancement of current technologies. We then systematically categorize the various families of high-resolution 3D histology techniques, discussing their core principles, capabilities, and inherent limitations. These 3D histology techniques include microscopy imaging, tomographic approaches, single-cell and spatial omics, computational methods and 3D tissue reconstruction (e.g. 3D cultures and spheroids). Additionally, we explore a wide range of applications for single-cell 3D histology, demonstrating how single-cell and spatial technologies are being utilized in the fields such as oncology, cardiology, neuroscience, immunology, developmental biology and regenerative medicine. Despite the remarkable progress made in recent years, the field still faces significant challenges, including high barriers to entry, issues with data robustness, ambiguous best practices for experimental design, and a lack of standardization across methodologies. This review offers a thorough analysis of these challenges and presents recommendations to surmount them, with the overarching goal of nurturing ongoing innovation and broader integration of cellular 3D tissue analysis in both biology research and clinical practice.

Decoding the blueprints of embryo development with single-cell and spatial omics

Embryonic development is a complex and intricately regulated process that encompasses precise control over cell differentiation, morphogenesis, and the underlying gene expression changes. Recent years have witnessed a remarkable acceleration in the development of single-cell and spatial omic technologies, enabling high-throughput profiling of transcriptomic and other multi-omic information at the individual cell level. These innovations offer fresh and multifaceted perspectives for investigating the intricate cellular and molecular mechanisms that govern embryonic development. In this review, we provide an in-depth exploration of the latest technical advancements in single-cell and spatial multi-omic methodologies and compile a systematic catalog of their applications in the field of embryonic development. We deconstruct the research strategies employed by recent studies that leverage single-cell sequencing techniques and underscore the unique advantages of spatial transcriptomics. Furthermore, we delve into both the current applications, data analysis algorithms and the untapped potential of these technologies in advancing our understanding of embryonic development. With the continuous evolution of multi-omic technologies, we anticipate their widespread adoption and profound contributions to unraveling the intricate molecular foundations underpinning embryo development in the foreseeable future.

Advanced Carriers for Precise Delivery and Therapeutic Mechanisms of Traditional Chinese Medicines: Integrating Spatial Multi-Omics and Delivery Visualization

The complex composition of traditional Chinese medicines (TCMs) has posed challenges for in-depth study and global application, despite their abundance of bioactive compounds that make them valuable resources for disease treatment. To overcome these obstacles, it is essential to modernize TCMs by focusing on precise disease treatment. This involves elucidating the structure-activity relationships within their complex compositions, ensuring accurate in vivo delivery, and monitoring the delivery process. This review discusses the research progress of TCMs in precision disease treatment from three perspectives: spatial multi-omics technology for precision therapeutic activity, carrier systems for precise in vivo delivery, and medical imaging technology for visualizing the delivery process. The aim is to establish a novel research paradigm that advances the precision therapy of TCMs.

Unveiling patterns in spatial transcriptomics data: a novel approach utilizing graph attention autoencoder and multiscale deep subspace clustering network

BACKGROUND

The accurate deciphering of spatial domains, along with the identification of differentially expressed genes and the inference of cellular trajectory based on spatial transcriptomic (ST) data, holds significant potential for enhancing our understanding of tissue organization and biological functions. However, most of spatial clustering methods can neither decipher complex structures in ST data nor entirely employ features embedded in different layers.

RESULTS

This article introduces STMSGAL, a novel framework for analyzing ST data by incorporating graph attention autoencoder and multiscale deep subspace clustering. First, STMSGAL constructs ctaSNN, a cell type-aware shared nearest neighbor graph, using Louvian clustering exclusively based on gene expression profiles. Subsequently, it integrates expression profiles and ctaSNN to generate spot latent representations using a graph attention autoencoder and multiscale deep subspace clustering. Lastly, STMSGAL implements spatial clustering, differential expression analysis, and trajectory inference, providing comprehensive capabilities for thorough data exploration and interpretation. STMSGAL was evaluated against 7 methods, including SCANPY, SEDR, CCST, DeepST, GraphST, STAGATE, and SiGra, using four 10x Genomics Visium datasets, 1 mouse visual cortex STARmap dataset, and 2 Stereo-seq mouse embryo datasets. The comparison showcased STMSGAL's remarkable performance across Davies-Bouldin, Calinski-Harabasz, S_Dbw, and ARI values. STMSGAL significantly enhanced the identification of layer structures across ST data with different spatial resolutions and accurately delineated spatial domains in 2 breast cancer tissues, adult mouse brain (FFPE), and mouse embryos.

CONCLUSIONS

STMSGAL can serve as an essential tool for bridging the analysis of cellular spatial organization and disease pathology, offering valuable insights for researchers in the field.

Spatial Transcriptomics: Biotechnologies, Computational Tools, and Neuroscience Applications

Spatial transcriptomics (ST) represents a revolutionary approach in molecular biology, providing unprecedented insights into the spatial organization of gene expression within tissues. This review aims to elucidate advancements in ST technologies, their computational tools, and their pivotal applications in neuroscience. It is begun with a historical overview, tracing the evolution from early image-based techniques to contemporary sequence-based methods. Subsequently, the computational methods essential for ST data analysis, including preprocessing, cell type annotation, spatial clustering, detection of spatially variable genes, cell-cell interaction analysis, and 3D multi-slices integration are discussed. The central focus of this review is the application of ST in neuroscience, where it has significantly contributed to understanding the brain's complexity. Through ST, researchers advance brain atlas projects, gain insights into brain development, and explore neuroimmune dysfunctions, particularly in brain tumors. Additionally, ST enhances understanding of neuronal vulnerability in neurodegenerative diseases like Alzheimer's and neuropsychiatric disorders such as schizophrenia. In conclusion, while ST has already profoundly impacted neuroscience, challenges remain issues such as enhancing sequencing technologies and developing robust computational tools. This review underscores the transformative potential of ST in neuroscience, paving the way for new therapeutic insights and advancements in brain research.

Computational Approaches in Spatial Transcriptomics for the Study of Mammalian Spermatogenesis

Spermatogenesis is a complex and dynamic cellular differentiation process critical to male fertility. Although the full continuum of gene expression patterns from spermatogonial stem cells (SSCs) to spermatozoa in steady state was characterized using single-cell RNA sequencing technologies, the transcriptional dynamics of spermatogenesis within its native tissue context was largely unexplored. The recent development of spatial transcriptomics (ST) technologies has transformed male fertility research from a single-cell level to a two-dimensional spatial coordinate system and facilitated the study of spermatogenesis in the native environment of both the rodent and human testes. The spatial gene expression information generated by these ST technologies requires new computational approaches to extract novel biological insights. These requirements include, but are not limited to, spatial mapping of testicular cell types, identifying spatially variable genes, and understanding the molecular cross-talk between testicular cell types. Here, we review computational approaches that have been used to dissect mammalian spermatogenesis in the context of ST. We also highlight new computational approaches that can be leveraged to reveal novel insights into male fertility.

Spatiotemporal modeling of molecular holograms

Quantifying spatiotemporal dynamics during embryogenesis is crucial for understanding congenital diseases. We developed Spateo (https://github.com/aristoteleo/spateo-release), a 3D spatiotemporal modeling framework, and applied it to a 3D mouse embryogenesis atlas at E9.5 and E11.5, capturing eight million cells. Spateo enables scalable, partial, non-rigid alignment, multi-slice refinement, and mesh correction to create molecular holograms of whole embryos. It introduces digitization methods to uncover multi-level biology from subcellular to whole organ, identifying expression gradients along orthogonal axes of emergent 3D structures, e.g., secondary organizers such as midbrain-hindbrain boundary (MHB). Spateo further jointly models intercellular and intracellular interaction to dissect signaling landscapes in 3D structures, including the zona limitans intrathalamica (ZLI). Lastly, Spateo introduces "morphometric vector fields" of cell migration and integrates spatial differential geometry to unveil molecular programs underlying asymmetrical murine heart organogenesis and others, bridging macroscopic changes with molecular dynamics. Thus, Spateo enables the study of organ ecology at a molecular level in 3D space over time.

FOXP Genes Regulate Purkinje Cell Diversity in Cerebellar Development and Evolution

Mammalian cerebellar development is thought to be influenced by distinct Purkinje cell (PC) subtypes. However, the degree of PC heterogeneity and the molecular drivers of this diversity have remained unclear, hindering efforts to manipulate PC diversification and assess its role in cerebellar development. Here, we demonstrate the critical role of Foxp genes in cerebellar development by regulating PC diversification. We identified 11 PC subtypes in the embryonic mouse cerebellum through single-cell RNA sequencing. Using a novel unsupervised method, we mapped these subtypes in three-dimensional space, revealing discrete PC subtypes predictive of adult cerebellar organization, including longitudinal stripes and lobules. These subtypes exhibit unique combinations of Foxp1 , Foxp2 , and Foxp4 expression. Deletion of Foxp2 and Foxp1 disrupts PC diversification, leading to altered cerebellar patterning, including the loss of a specific Foxp1 -expressing subtype and the cerebellar hemisphere. The Foxp1 -expressing PC subtype is much more abundant in the fetal human cerebellum than in mice, but rare in the chick cerebellum, correlating with cerebellar hemisphere size in these species. This highlights the significance of Foxp1 -expressing PCs in cerebellar hemisphere development and evolution. Therefore, our study identifies Foxp genes as key regulators of PC diversity, providing new insights into the developmental and evolutionary underpinnings of the cerebellum.

Graph domain adaptation-based framework for gene expression enhancement and cell type identification in large-scale spatially resolved transcriptomics

Spatially resolved transcriptomics (SRT) technologies facilitate gene expression profiling with spatial resolution in a naïve state. Nevertheless, current SRT technologies exhibit limitations, manifesting as either low transcript detection sensitivity or restricted gene throughput. These constraints result in diminished precision and coverage in gene measurement. In response, we introduce SpaGDA, a sophisticated deep learning-based graph domain adaptation framework for both scenarios of gene expression imputation and cell type identification in spatially resolved transcriptomics data by impartially transferring knowledge from reference scRNA-seq data. Systematic benchmarking analyses across several SRT datasets generated from different technologies have demonstrated SpaGDA's superior effectiveness compared to state-of-the-art methods in both scenarios. Further applied to three SRT datasets of different biological contexts, SpaGDA not only better recovers the well-established knowledge sourced from public atlases and existing scientific literature but also yields a more informative spatial expression pattern of genes. Together, these results demonstrate that SpaGDA can be used to overcome the challenges of current SRT data and provide more accurate insights into biological processes or disease development. The SpaGDA is available in https://github.com/shenrb/SpaGDA.

Multimodal contrastive learning for spatial gene expression prediction using histology images

In recent years, the advent of spatial transcriptomics (ST) technology has unlocked unprecedented opportunities for delving into the complexities of gene expression patterns within intricate biological systems. Despite its transformative potential, the prohibitive cost of ST technology remains a significant barrier to its widespread adoption in large-scale studies. An alternative, more cost-effective strategy involves employing artificial intelligence to predict gene expression levels using readily accessible whole-slide images stained with Hematoxylin and Eosin (H&E). However, existing methods have yet to fully capitalize on multimodal information provided by H&E images and ST data with spatial location. In this paper, we propose mclSTExp, a multimodal contrastive learning with Transformer and Densenet-121 encoder for Spatial Transcriptomics Expression prediction. We conceptualize each spot as a "word", integrating its intrinsic features with spatial context through the self-attention mechanism of a Transformer encoder. This integration is further enriched by incorporating image features via contrastive learning, thereby enhancing the predictive capability of our model. We conducted an extensive evaluation of highly variable genes in two breast cancer datasets and a skin squamous cell carcinoma dataset, and the results demonstrate that mclSTExp exhibits superior performance in predicting spatial gene expression. Moreover, mclSTExp has shown promise in interpreting cancer-specific overexpressed genes, elucidating immune-related genes, and identifying specialized spatial domains annotated by pathologists. Our source code is available at https://github.com/shizhiceng/mclSTExp.

Spatial landscapes of cancers: insights and opportunities

Solid tumours comprise many different cell types organized in spatially structured arrangements, with substantial intratumour and intertumour heterogeneity. Advances in spatial profiling technologies over the past decade hold promise to capture the complexity of these cellular architectures to build a holistic view of the intricate molecular mechanisms that shape the tumour ecosystem. Some of these mechanisms act at the cellular scale and are controlled by cell-autonomous programmes or communication between nearby cells, whereas other mechanisms result from coordinated efforts between large networks of cells and extracellular molecules organized into tissues and organs. In this Review we provide insights into the application of single-cell and spatial profiling tools, with a focus on spatially resolved transcriptomic tools developed to understand the cellular architecture of the tumour microenvironment and identify opportunities to use them to improve clinical management of cancers.

Spatiotemporal omics for biology and medicine

The completion of the Human Genome Project has provided a foundational blueprint for understanding human life. Nonetheless, understanding the intricate mechanisms through which our genetic blueprint is involved in disease or orchestrates development across temporal and spatial dimensions remains a profound scientific challenge. Recent breakthroughs in cellular omics technologies have paved new pathways for understanding the regulation of genomic elements and the relationship between gene expression, cellular functions, and cell fate determination. The advent of spatial omics technologies, encompassing both imaging and sequencing-based methodologies, has enabled a comprehensive understanding of biological processes from a cellular ecosystem perspective. This review offers an updated overview of how spatial omics has advanced our understanding of the translation of genetic information into cellular heterogeneity and tissue structural organization and their dynamic changes over time. It emphasizes the discovery of various biological phenomena, related to organ functionality, embryogenesis, species evolution, and the pathogenesis of diseases.

Benchmarking clustering, alignment, and integration methods for spatial transcriptomics

BACKGROUND

Spatial transcriptomics (ST) is advancing our understanding of complex tissues and organisms. However, building a robust clustering algorithm to define spatially coherent regions in a single tissue slice and aligning or integrating multiple tissue slices originating from diverse sources for essential downstream analyses remains challenging. Numerous clustering, alignment, and integration methods have been specifically designed for ST data by leveraging its spatial information. The absence of comprehensive benchmark studies complicates the selection of methods and future method development.

RESULTS

In this study, we systematically benchmark a variety of state-of-the-art algorithms with a wide range of real and simulated datasets of varying sizes, technologies, species, and complexity. We analyze the strengths and weaknesses of each method using diverse quantitative and qualitative metrics and analyses, including eight metrics for spatial clustering accuracy and contiguity, uniform manifold approximation and projection visualization, layer-wise and spot-to-spot alignment accuracy, and 3D reconstruction, which are designed to assess method performance as well as data quality. The code used for evaluation is available on our GitHub. Additionally, we provide online notebook tutorials and documentation to facilitate the reproduction of all benchmarking results and to support the study of new methods and new datasets.

CONCLUSIONS

Our analyses lead to comprehensive recommendations that cover multiple aspects, helping users to select optimal tools for their specific needs and guide future method development.

SANTO: a coarse-to-fine alignment and stitching method for spatial omics

With the flourishing of spatial omics technologies, alignment and stitching of slices becomes indispensable to decipher a holistic view of 3D molecular profile. However, existing alignment and stitching methods are unpractical to process large-scale and image-based spatial omics dataset due to extreme time consumption and unsatisfactory accuracy. Here we propose SANTO, a coarse-to-fine method targeting alignment and stitching tasks for spatial omics. SANTO firstly rapidly supplies reasonable spatial positions of two slices and identifies the overlap region. Then, SANTO refines the positions of two slices by considering spatial and omics patterns. Comprehensive experiments demonstrate the superior performance of SANTO over existing methods. Specifically, SANTO stitches cross-platform slices for breast cancer samples, enabling integration of complementary features to synergistically explore tumor microenvironment. SANTO is then applied to 3D-to-3D spatiotemporal alignment to study development of mouse embryo. Furthermore, SANTO enables cross-modality alignment of spatial transcriptomic and epigenomic data to understand complementary interactions.

Data enhancement in the age of spatial biology

Unveiling the intricate interplay of cells in their native environment lies at the heart of understanding fundamental biological processes and unraveling disease mechanisms, particularly in complex diseases like cancer. Spatial transcriptomics (ST) offers a revolutionary lens into the spatial organization of gene expression within tissues, empowering researchers to study both cell heterogeneity and microenvironments in health and disease. However, current ST technologies often face limitations in either resolution or the number of genes profiled simultaneously. Integrating ST data with complementary sources, such as single-cell transcriptomics and detailed tissue staining images, presents a powerful solution to overcome these limitations. This review delves into the computational approaches driving the integration of spatial transcriptomics with other data types. By illuminating the key challenges and outlining the current algorithmic solutions, we aim to highlight the immense potential of these methods to revolutionize our understanding of cancer biology.

HLAIImaster: a deep learning method with adaptive domain knowledge predicts HLA II neoepitope immunogenic responses

While significant strides have been made in predicting neoepitopes that trigger autologous CD4+ T cell responses, accurately identifying the antigen presentation by human leukocyte antigen (HLA) class II molecules remains a challenge. This identification is critical for developing vaccines and cancer immunotherapies. Current prediction methods are limited, primarily due to a lack of high-quality training epitope datasets and algorithmic constraints. To predict the exogenous HLA class II-restricted peptides across most of the human population, we utilized the mass spectrometry data to profile >223 000 eluted ligands over HLA-DR, -DQ, and -DP alleles. Here, by integrating these data with peptide processing and gene expression, we introduce HLAIImaster, an attention-based deep learning framework with adaptive domain knowledge for predicting neoepitope immunogenicity. Leveraging diverse biological characteristics and our enhanced deep learning framework, HLAIImaster is significantly improved against existing tools in terms of positive predictive value across various neoantigen studies. Robust domain knowledge learning accurately identifies neoepitope immunogenicity, bridging the gap between neoantigen biology and the clinical setting and paving the way for future neoantigen-based therapies to provide greater clinical benefit. In summary, we present a comprehensive exploitation of the immunogenic neoepitope repertoire of cancers, facilitating the effective development of "just-in-time" personalized vaccines.

Navigating the landscapes of spatial transcriptomics: How computational methods guide the way

Spatially resolved transcriptomics has been dramatically transforming biological and medical research in various fields. It enables transcriptome profiling at single-cell, multi-cellular, or sub-cellular resolution, while retaining the information of geometric localizations of cells in complex tissues. The coupling of cell spatial information and its molecular characteristics generates a novel multi-modal high-throughput data source, which poses new challenges for the development of analytical methods for data-mining. Spatial transcriptomic data are often highly complex, noisy, and biased, presenting a series of difficulties, many unresolved, for data analysis and generation of biological insights. In addition, to keep pace with the ever-evolving spatial transcriptomic experimental technologies, the existing analytical theories and tools need to be updated and reformed accordingly. In this review, we provide an overview and discussion of the current computational approaches for mining of spatial transcriptomics data. Future directions and perspectives of methodology design are proposed to stimulate further discussions and advances in new analytical models and algorithms. This article is categorized under: RNA Methods > RNA Analyses in Cells RNA Evolution and Genomics > Computational Analyses of RNA RNA Export and Localization > RNA Localization.

Spatially Resolved Multiomics: Data Analysis from Monoomics to Multiomics

Spatial monoomics has been recognized as a powerful tool for exploring life sciences. Recently, spatial multiomics has advanced considerably, which could contribute to clarifying many biological issues. Spatial monoomics techniques in epigenomics, genomics, transcriptomics, proteomics, and metabolomics can enhance our understanding of biological functions and cellular identities by simultaneously measuring tissue structures and biomolecule levels. Spatial monoomics technology has evolved from monoomics to spatial multiomics. Moreover, the spatial resolution, high-throughput detection capability, capture efficiency, and compatibility with various sample types of omics technology have considerably advanced. Despite the technological advances in this field, data analysis frameworks have stagnated. Current challenges include incomplete spatial monoomics data analysis pipeline, overly complex data analysis tasks, and few established spatial multiomics data analysis strategies. In this review, we systematically summarize recent developments of various spatial monoomics techniques and improvements in related data analysis pipeline. On the basis of the spatial multiomics technology, we propose a data integration strategy with cross-platform, cross-slice, and cross-modality. We summarize the potential applications of spatial monoomics technology, aiming to provide researchers and clinicians with a better understanding of how such applications have advanced. Spatial multiomics technology is expected to substantially impact biology and precision medicine through measurements of cellular tissue structures and the extraction of biomolecular features.