Label-free metabolic optical biomarkers track stem cell fate transition in real time.

Tracking stem cell fate transition is crucial for understanding their development and optimizing biomanufacturing. Destructive single-cell methods provide a pseudotemporal landscape of stem cell differentiation but cannot monitor stem cell fate in real time. We established a metabolic optical metric using label-free fluorescence lifetime imaging microscopy (FLIM), feature extraction and machine learning-assisted analysis, for real-time cell fate tracking. From a library of 205 metabolic optical biomarker (MOB) features, we identified 56 associated with hematopoietic stem cell (HSC) differentiation. These features collectively describe HSC fate transition and detect its bifurcate lineage choice. We further derived a MOB score measuring the "metabolic stemness" of single cells and distinguishing their division patterns. This score reveals a distinct role of asymmetric division in rescuing stem cells with compromised metabolic stemness and a unique mechanism of PI3K inhibition in promoting ex vivo HSC maintenance. MOB profiling is a powerful tool for tracking stem cell fate transition and improving their biomanufacturing from a single-cell perspective.

A new dawn for the study of cell type evolution.

Animal evolution is influenced by the emergence of new cell types, yet our understanding of this process remains elusive. This prompts the need for a broader exploration across diverse research organisms, facilitated by recent breakthroughs, such as gene editing tools and single-cell genomics. Essential to our understanding of cell type evolution is the accurate identification of homologous cells. We delve into the significance of considering developmental ontogeny and potential pitfalls when drawing conclusions about cell type homology. Additionally, we highlight recent discoveries in the study of cell type evolution through the application of single-cell transcriptomics and pinpoint areas ripe for further exploration.

GRouNdGAN: GRN-guided simulation of single-cell RNA-seq data using causal generative adversarial networks.

We introduce GRouNdGAN, a gene regulatory network (GRN)-guided reference-based causal implicit generative model for simulating single-cell RNA-seq data, in silico perturbation experiments, and benchmarking GRN inference methods. Through the imposition of a user-defined GRN in its architecture, GRouNdGAN simulates steady-state and transient-state single-cell datasets where genes are causally expressed under the control of their regulating transcription factors (TFs). Training on six experimental reference datasets, we show that our model captures non-linear TF-gene dependencies and preserves gene identities, cell trajectories, pseudo-time ordering, and technical and biological noise, with no user manipulation and only implicit parameterization. GRouNdGAN can synthesize cells under new conditions to perform in silico TF knockout experiments. Benchmarking various GRN inference algorithms reveals that GRouNdGAN effectively bridges the existing gap between simulated and biological data benchmarks of GRN inference algorithms, providing gold standard ground truth GRNs and realistic cells corresponding to the biological system of interest.

Screening and validation of atherosclerosis PAN-apoptotic immune-related genes based on single-cell sequencing.

Background: Carotid atherosclerosis (CAS) is a complication of atherosclerosis (AS). PAN-optosome is an inflammatory programmed cell death pathway event regulated by the PAN-optosome complex. CAS's PAN-optosome-related genes (PORGs) have yet to be studied. Hence, screening the PAN-optosome-related diagnostic genes for treating CAS was vital. Methods: We introduced transcriptome data to screen out differentially expressed genes (DEGs) in CAS. Subsequently, WGCNA analysis was utilized to mine module genes about PANoptosis score. We performed differential expression analysis (CAS samples vs. standard samples) to obtain CAS-related differentially expressed genes at the single-cell level. Venn diagram was executed to identify PAN-optosome-related differential genes (POR-DEGs) associated with CAS. Further, LASSO regression and RF algorithm were implemented to were executed to build a diagnostic model. We additionally performed immune infiltration and gene set enrichment analysis (GSEA) based on diagnostic genes. We verified the accuracy of the model genes by single-cell nuclear sequencing and RT-qPCR validation of clinical samples, as well as in vitro cellular experiments. Results: We identified 785 DEGs associated with CAS. Then, 4296 module genes about PANoptosis score were obtained. We obtained the 7365 and 1631 CAS-related DEGs at the single-cell level, respectively. 67 POR-DEGs were retained Venn diagram. Subsequently, 4 PAN-optosome-related diagnostic genes (CNTN4, FILIP1, PHGDH, and TFPI2) were identified via machine learning. Cellular function tests on four genes showed that these genes have essential roles in maintaining arterial cell viability and resisting cellular senescence. Conclusion: We obtained four PANoptosis-related diagnostic genes (CNTN4, FILIP1, PHGDH, and TFPI2) associated with CAS, laying a theoretical foundation for treating CAS.

Diagnostic leukapheresis reveals distinct phenotypes of NSCLC circulating tumor cells.

BACKGROUND: Circulating tumor cells (CTCs) hold immense promise for unraveling tumor heterogeneity and understanding treatment resistance. However, conventional methods, especially in cancers like non-small cell lung cancer (NSCLC), often yield low CTC numbers, hindering comprehensive analyses. This study addresses this limitation by employing diagnostic leukapheresis (DLA) to cancer patients, enabling the screening of larger blood volumes. To leverage DLA's full potential, this study introduces a novel approach for CTC enrichment from DLAs. METHODS: DLA was applied to six advanced stage NSCLC patients. For an unbiased CTC enrichment, a two-step approach based on negative depletion of hematopoietic cells was used. Single-cell (sc) whole-transcriptome sequencing was performed, and CTCs were identified based on gene signatures and inferred copy number variations. RESULTS: Remarkably, this innovative approach led to the identification of unprecedented 3,363 CTC transcriptomes. The extensive heterogeneity among CTCs was unveiled, highlighting distinct phenotypes related to the epithelial-mesenchymal transition (EMT) axis, stemness, immune responsiveness, and metabolism. Comparison with sc transcriptomes from primary NSCLC cells revealed that CTCs encapsulate the heterogeneity of their primary counterparts while maintaining unique CTC-specific phenotypes. CONCLUSIONS: In conclusion, this study pioneers a transformative method for enriching CTCs from DLA, resulting in a substantial increase in CTC numbers. This allowed the creation of the first-ever single-cell whole transcriptome in-depth characterization of the heterogeneity of over 3,300 NSCLC-CTCs. The findings not only confirm the diagnostic value of CTCs in monitoring tumor heterogeneity but also propose a CTC-specific signature that can be exploited for targeted CTC-directed therapies in the future. This comprehensive approach signifies a major leap forward, positioning CTCs as a key player in advancing our understanding of cancer dynamics and paving the way for tailored therapeutic interventions.

Comparative single-cell multiplex immunophenotyping of therapy-naive patients with rheumatoid arthritis, systemic sclerosis, and systemic lupus erythematosus shed light on disease-specific composition of the peripheral immune system.

Introduction: Systemic autoimmune diseases (SADs) are a significant burden on the healthcare system. Understanding the complexity of the peripheral immunophenotype in SADs may facilitate the differential diagnosis and identification of potential therapeutic targets. Methods: Single-cell mass cytometric immunophenotyping was performed on peripheral blood mononuclear cells (PBMCs) from healthy controls (HCs) and therapy-naive patients with rheumatoid arthritis (RA), progressive systemic sclerosis (SSc), and systemic lupus erythematosus (SLE). Immunophenotyping was performed on 15,387,165 CD45+ live single cells from 52 participants (13 cases/group), using an antibody panel to detect 34 markers. Results: Using the t-SNE (t-distributed stochastic neighbor embedding) algorithm, the following 17 main immune cell types were determined: CD4+/CD57- T cells, CD4+/CD57+ T cells, CD8+/CD161- T cells, CD8+/CD161+/CD28+ T cells, CD8dim T cells, CD3+/CD4-/CD8- T cells, TCRγ/δ T cells, CD4+ NKT cells, CD8+ NKT cells, classic NK cells, CD56dim/CD98dim cells, B cells, plasmablasts, monocytes, CD11cdim/CD172dim cells, myeloid dendritic cells (mDCs), and plasmacytoid dendritic cells (pDCs). Seven of the 17 main cell types exhibited statistically significant frequencies in the investigated groups. The expression levels of the 34 markers in the main populations were compared between HCs and SADs. In summary, 59 scatter plots showed significant differences in the expression intensities between at least two groups. Next, each immune cell population was divided into subpopulations (metaclusters) using the FlowSOM (self-organizing map) algorithm. Finally, 121 metaclusters (MCs) of the 10 main immune cell populations were found to have significant differences to classify diseases. The single-cell T-cell heterogeneity represented 64MCs based on the expression of 34 markers, and the frequency of 23 MCs differed significantly between at least twoconditions. The CD3- non-T-cell compartment contained 57 MCs with 17 MCs differentiating at least two investigated groups. In summary, we are the first to demonstrate the complexity of the immunophenotype of 34 markers over 15 million single cells in HCs vs. therapy-naive patients with RA, SSc, and SLE. Disease specific population frequencies or expression patterns of peripheral immune cells provide a single-cell data resource to the scientific community.

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

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

Identification and validation of a prognostic anoikis-related gene signature in papillary thyroid carcinoma by integrated analysis of single-cell and bulk RNA-sequencing.

Papillary thyroid carcinoma (PTC) prognosis may be deteriorated due to the metastases, and anoikis palys an essential role in the tumor metastasis. However, the potential effect of anoikis-related genes on the prognosis of PTC was unclear. The mRNA and clinical information were obtained from the cancer genome atlas database. Hub genes were identified and risk model was constructed using Cox regression analysis. Kaplan-Meier (K-M) curve was applied for the survival analysis. Immune infiltration and immune therapy response were calculated using CIBERSORT and TIDE. The identification of cell types and cell interaction was performed by Seurat, SingleR and CellChat packages. GO, KEGG, and GSVA were applied for the enrichment analysis. Protein-protein interaction network was constructed in STRING and Cytoscape. Drug sensitivity was assessed in GSCA. Based on bulk RNA data, we identified 4 anoikis-related risk signatures, which were oncogenes, and constructed a risk model. The enrichment analysis found high risk group was enriched in some immune-related pathways. High risk group had higher infiltration of Tregs, higher TIDE score and lower levels of monocytes and CD8 T cells. Based on scRNA data, we found that 4 hub genes were mainly expressed in monocytes and macrophages, and they interacted with T cells. Hub genes were significantly related to immune escape-related genes. Drug sensitivity analysis suggested that cyclin dependent kinase inhibitor 2A may be a better chemotherapy target. We constructed a risk model which could effectively and steadily predict the prognosis of PTC. We inferred that the immune escape may be involved in the development of PTC.

Single-cell RNA-seq data analysis reveals functionally relevant biomarkers of early brain development and their regulatory footprints in human embryonic stem cells (hESCs).

The complicated process of neuronal development is initiated early in life, with the genetic mechanisms governing this process yet to be fully elucidated. Single-cell RNA sequencing (scRNA-seq) is a potent instrument for pinpointing biomarkers that exhibit differential expression across various cell types and developmental stages. By employing scRNA-seq on human embryonic stem cells, we aim to identify differentially expressed genes (DEGs) crucial for early-stage neuronal development. Our focus extends beyond simply identifying DEGs. We strive to investigate the functional roles of these genes through enrichment analysis and construct gene regulatory networks to understand their interactions. Ultimately, this comprehensive approach aspires to illuminate the molecular mechanisms and transcriptional dynamics governing early human brain development. By uncovering potential links between these DEGs and intelligence, mental disorders, and neurodevelopmental disorders, we hope to shed light on human neurological health and disease. In this study, we have used scRNA-seq to identify DEGs involved in early-stage neuronal development in hESCs. The scRNA-seq data, collected on days 26 (D26) and 54 (D54), of the in vitro differentiation of hESCs to neurons were analyzed. Our analysis identified 539 DEGs between D26 and D54. Functional enrichment of those DEG biomarkers indicated that the up-regulated DEGs participated in neurogenesis, while the down-regulated DEGs were linked to synapse regulation. The Reactome pathway analysis revealed that down-regulated DEGs were involved in the interactions between proteins located in synapse pathways. We also discovered interactions between DEGs and miRNA, transcriptional factors (TFs) and DEGs, and between TF and miRNA. Our study identified 20 significant transcription factors, shedding light on early brain development genetics. The identified DEGs and gene regulatory networks are valuable resources for future research into human brain development and neurodevelopmental disorders.

Coupling imaging mass cytometry with Alcian blue histochemical staining for a single-slide approach.

Imaging mass cytometry (IMC) is a metal mass spectrometry-based method allowing highly multiplex immunophenotyping of cells within tissue samples. However, some limitations of IMC are its 1-µm resolution and its time and costs of analysis limiting respectively the detailed histopathological analysis of IMC-produced images and its application to small selected tissue regions of interest (ROI) of one to few square millimeters. Coupling on a single-tissue section, IMC and histopathological analyses could permit a better selection of the ROI for IMC analysis as well as co-analysis of immunophenotyping and histopathological data until the single-cell level. The development of this method is the aim of the present study in which we point to the feasibility of applying the IMC process to tissue sections previously Alcian blue-stained and digitalized before IMC tissue destructive analyses. This method could help to improve the process of IMC in terms of ROI selection, time of analysis, and the confrontation between histopathological and immunophenotypic data of cells.

EmptyDropsMultiome discriminates real cells from background in single-cell multiomics assays.

Multiomic droplet-based technologies allow different molecular modalities, such as chromatin accessibility and gene expression (scATAC-seq and scRNA-seq), to be probed in the same nucleus. We develop EmptyDropsMultiome, an approach that distinguishes true nuclei-containing droplets from background. Using simulations, we show that EmptyDropsMultiome has higher statistical power and accuracy than existing approaches, including CellRanger-arc and EmptyDrops. On real datasets, we observe that CellRanger-arc misses more than half of the nuclei identified by EmptyDropsMultiome and, moreover, is biased against certain cell types, some of which have a retrieval rate lower than 20%.

New genetic insights into immunotherapy outcomes in gastric cancer via single-cell RNA sequencing and random forest model.

OBJECTIVE: The high mortality rate of gastric cancer, traditionally managed through surgery, underscores the urgent need for advanced therapeutic strategies. Despite advancements in treatment modalities, outcomes remain suboptimal, necessitating the identification of novel biomarkers to predict sensitivity to immunotherapy. This study focuses on utilizing single-cell sequencing for gene identification and developing a random forest model to predict immunotherapy sensitivity in gastric cancer patients. METHODS: Differentially expressed genes were identified using single-cell RNA sequencing (scRNA-seq) and gene set enrichment analysis (GESA). A random forest model was constructed based on these genes, and its effectiveness was validated through prognostic analysis. Further, analyses of immune cell infiltration, immune checkpoints, and the random forest model provided deeper insights. RESULTS: High METTL1 expression was found to correlate with improved survival rates in gastric cancer patients (P = 0.042), and the random forest model, based on METTL1 and associated prognostic genes, achieved a significant predictive performance (AUC = 0.863). It showed associations with various immune cell types and negative correlations with CTLA4 and PDCD1 immune checkpoints. Experiments in vitro and in vivo demonstrated that METTL1 enhances gastric cancer cell activity by suppressing T cell proliferation and upregulating CTLA4 and PDCD1. CONCLUSION: The random forest model, based on scRNA-seq, shows high predictive value for survival and immunotherapy sensitivity in gastric cancer patients. This study underscores the potential of METTL1 as a biomarker in enhancing the efficacy of gastric cancer immunotherapy.

Molecular mechanism of radiation tolerance in lung adenocarcinoma cells using single-cell RNA sequencing.

The efficacy of radiotherapy, a cornerstone in the treatment of lung adenocarcinoma (LUAD), is profoundly undermined by radiotolerance. This resistance not only poses a significant clinical challenge but also compromises patient survival rates. Therefore, it is important to explore this mechanism for the treatment of LUAD. Multiple public databases were used for single-cell RNA sequencing (scRNA-seq) data. We filtered, normalized and downscaled scRNA-seq data based on the Seurat package to obtain different cell subpopulations. Subsequently, the ssGSEA algorithm was used to assess the enrichment scores of the different cell subpopulations, and thus screen the cell subpopulations that are most relevant to radiotherapy tolerance based on the Pearson method. Finally, pseudotime analysis was performed, and a preliminary exploration of gene mutations in different cell subpopulations was performed. We identified HIST1H1D+ A549 and PIF1+ A549 as the cell subpopulations related to radiotolerance. The expression levels of cell cycle-related genes and pathway enrichment scores of these two cell subpopulations increased gradually with the extension of radiation treatment time. Finally, we found that the proportion of TP53 mutations in patients who had received radiotherapy was significantly higher than that in patients who had not received radiotherapy. We identified two cellular subpopulations associated with radiotherapy tolerance, which may shed light on the molecular mechanisms of radiotherapy tolerance in LUAD and provide new clinical perspectives.

Isolation of Nuclei from Human Intermuscular Adipose Tissue and Downstream Single-Nuclei RNA Sequencing.

Intermuscular adipose tissue (IMAT) is a relatively understudied adipose depot located between muscle fibers. IMAT content increases with age and BMI and is associated with metabolic and muscle degenerative diseases; however, an understanding of the biological properties of IMAT and its interplay with the surrounding muscle fibers is severely lacking. In recent years, single-cell and nuclei RNA sequencing have provided us with cell type-specific atlases of several human tissues. However, the cellular composition of human IMAT remains largely unexplored due to the inherent challenges of its accessibility from biopsy collection in humans. In addition to the limited amount of tissue collected, the processing of human IMAT is complicated due to its proximity to skeletal muscle tissue and fascia. The lipid-laden nature of the adipocytes makes it incompatible with single-cell isolation. Hence, single nuclei RNA sequencing is optimal for obtaining high-dimensional transcriptomics at single-cell resolution and provides the potential to uncover the biology of this depot, including the exact cellular composition of IMAT. Here, we present a detailed protocol for nuclei isolation and library preparation of frozen human IMAT for single nuclei RNA sequencing. This protocol allows for the profiling of thousands of nuclei using a droplet-based approach, thus providing the capacity to detect rare and low-abundant cell types.

Reference-free inferring of transcriptomic events in cancer cells on single-cell data.

BACKGROUND: Cancerous cells' identity is determined via a mixture of multiple factors such as genomic variations, epigenetics, and the regulatory variations that are involved in transcription. The differences in transcriptome expression as well as abnormal structures in peptides determine phenotypical differences. Thus, bulk RNA-seq and more recent single-cell RNA-seq data (scRNA-seq) are important to identify pathogenic differences. In this case, we rely on k-mer decomposition of sequences to identify pathogenic variations in detail which does not need a reference, so it outperforms more traditional Next-Generation Sequencing (NGS) analysis techniques depending on the alignment of the sequences to a reference. RESULTS: Via our alignment-free analysis, over esophageal and glioblastoma cancer patients, high-frequency variations over multiple different locations (repeats, intergenic regions, exons, introns) as well as multiple different forms (fusion, polyadenylation, splicing, etc.) could be discovered. Additionally, we have analyzed the importance of less-focused events systematically in a classic transcriptome analysis pipeline where these events are considered as indicators for tumor prognosis, tumor prediction, tumor neoantigen inference, as well as their connection with respect to the immune microenvironment. CONCLUSIONS: Our results suggest that esophageal cancer (ESCA) and glioblastoma processes can be explained via pathogenic microbial RNA, repeated sequences, novel splicing variants, and long intergenic non-coding RNAs (lincRNAs). We expect our application of reference-free process and analysis to be helpful in tumor and normal samples differential scRNA-seq analysis, which in turn offers a more comprehensive scheme for major cancer-associated events.

Error modelled gene expression analysis (EMOGEA) provides a superior overview of time course RNA-seq measurements and low count gene expression.

Temporal RNA-sequencing (RNA-seq) studies of bulk samples provide an opportunity for improved understanding of gene regulation during dynamic phenomena such as development, tumor progression or response to an incremental dose of a pharmacotherapeutic. Moreover, single-cell RNA-seq (scRNA-seq) data implicitly exhibit temporal characteristics because gene expression values recapitulate dynamic processes such as cellular transitions. Unfortunately, temporal RNA-seq data continue to be analyzed by methods that ignore this ordinal structure and yield results that are often difficult to interpret. Here, we present Error Modelled Gene Expression Analysis (EMOGEA), a framework for analyzing RNA-seq data that incorporates measurement uncertainty, while introducing a special formulation for those acquired to monitor dynamic phenomena. This method is specifically suited for RNA-seq studies in which low-count transcripts with small-fold changes lead to significant biological effects. Such transcripts include genes involved in signaling and non-coding RNAs that inherently exhibit low levels of expression. Using simulation studies, we show that this framework down-weights samples that exhibit extreme responses such as batch effects allowing them to be modeled with the rest of the samples and maintain the degrees of freedom originally envisioned for a study. Using temporal experimental data, we demonstrate the framework by extracting a cascade of gene expression waves from a well-designed RNA-seq study of zebrafish embryogenesis and an scRNA-seq study of mouse pre-implantation and provide unique biological insights into the regulation of genes in each wave. For non-ordinal measurements, we show that EMOGEA has a much higher rate of true positive calls and a vanishingly small rate of false negative discoveries compared to common approaches. Finally, we provide two packages in Python and R that are self-contained and easy to use, including test data.

Unveiling the Dynamics behind Glioblastoma Multiforme Single-Cell Data Heterogeneity.

Glioblastoma Multiforme is a brain tumor distinguished by its aggressiveness. We suggested that this aggressiveness leads single-cell RNA-sequence data (scRNA-seq) to span a representative portion of the cancer attractors domain. This conjecture allowed us to interpret the scRNA-seq heterogeneity as reflecting a representative trajectory within the attractor's domain. We considered factors such as genomic instability to characterize the cancer dynamics through stochastic fixed points. The fixed points were derived from centroids obtained through various clustering methods to verify our method sensitivity. This methodological foundation is based upon sample and time average equivalence, assigning an interpretative value to the data cluster centroids and supporting parameters estimation. We used stochastic simulations to reproduce the dynamics, and our results showed an alignment between experimental and simulated dataset centroids. We also computed the Waddington landscape, which provided a visual framework for validating the centroids and standard deviations as characterizations of cancer attractors. Additionally, we examined the stability and transitions between attractors and revealed a potential interplay between subtypes. These transitions might be related to cancer recurrence and progression, connecting the molecular mechanisms of cancer heterogeneity with statistical properties of gene expression dynamics. Our work advances the modeling of gene expression dynamics and paves the way for personalized therapeutic interventions.

Single-cell transcriptome analysis reveals heterogeneity of neutrophils in non-small cell lung cancer.

BACKGROUND: Lung cancer stands out as a highly perilous malignant tumor with severe implications for human health. There has been a growing interest in neutrophils as a result of their role in promoting cancer in recent years. Thus, the present study aimed to investigate the heterogeneity of neutrophils in non-small cell lung cancer (NSCLC). METHODS: Single-cell RNA sequencing of tumor-associated neutrophils (TANs) and polymorphonuclear neutrophils sourced from the Gene Expression Omnibus database was analyzed. Moreover, cell-cell communication, differentiation trajectories and transcription factor analyses were performed. RESULTS: Neutrophils were found to be closely associated with macrophages. Four major types of TANs were identified: a transitional subcluster that migrated from blood to tumor microenvironment (TAN-0), an inflammatory subcluster (TAN-1), a subpopulation that displayed a distinctive transcriptional signature (TAN-2) and a final differentiation state that promoted tumor formation (TAN-3). Meanwhile, TAN-3 displayed a marked increase in glycolytic activity. Finally, transcription factors were analyzed to uncover distinct TAN cluster-specific regulons. CONCLUSIONS: The discovery of the dynamic characteristics of TANs in the present study is anticipated to contribute to yielding a better understanding of the tumor microenvironment and advancing the treatment of NSCLC.

SpatialCells: automated profiling of tumor microenvironments with spatially resolved multiplexed single-cell data.

Cancer is a complex cellular ecosystem where malignant cells coexist and interact with immune, stromal and other cells within the tumor microenvironment (TME). Recent technological advancements in spatially resolved multiplexed imaging at single-cell resolution have led to the generation of large-scale and high-dimensional datasets from biological specimens. This underscores the necessity for automated methodologies that can effectively characterize molecular, cellular and spatial properties of TMEs for various malignancies. This study introduces SpatialCells, an open-source software package designed for region-based exploratory analysis and comprehensive characterization of TMEs using multiplexed single-cell data. The source code and tutorials are available at https://semenovlab.github.io/SpatialCells. SpatialCells efficiently streamlines the automated extraction of features from multiplexed single-cell data and can process samples containing millions of cells. Thus, SpatialCells facilitates subsequent association analyses and machine learning predictions, making it an essential tool in advancing our understanding of tumor growth, invasion and metastasis.

AnnoSpat annotates cell types and quantifies cellular arrangements from spatial proteomics.

Cellular composition and anatomical organization influence normal and aberrant organ functions. Emerging spatial single-cell proteomic assays such as Image Mass Cytometry (IMC) and Co-Detection by Indexing (CODEX) have facilitated the study of cellular composition and organization by enabling high-throughput measurement of cells and their localization directly in intact tissues. However, annotation of cell types and quantification of their relative localization in tissues remain challenging. To address these unmet needs for atlas-scale datasets like Human Pancreas Analysis Program (HPAP), we develop AnnoSpat (Annotator and Spatial Pattern Finder) that uses neural network and point process algorithms to automatically identify cell types and quantify cell-cell proximity relationships. Our study of data from IMC and CODEX shows the higher performance of AnnoSpat in rapid and accurate annotation of cell types compared to alternative approaches. Moreover, the application of AnnoSpat to type 1 diabetic, non-diabetic autoantibody-positive, and non-diabetic organ donor cohorts recapitulates known islet pathobiology and shows differential dynamics of pancreatic polypeptide (PP) cell abundance and CD8+ T cells infiltration in islets during type 1 diabetes progression.