Sci Adv. 2023 Sep 29;9(39):eadh4119. doi: 10.1126/sciadv.adh4119. Epub 2023 Sep 27.

ABSTRACT

Understanding cell state transitions and purposefully controlling them to improve therapies is a longstanding challenge in biological research and medicine. Here, we identify a transcriptional signature that distinguishes activated macrophages from the tuberculosis (TB) susceptible and resistant mice. We then apply the cSTAR (cell state transition assessment and regulation) approach to data from screening-by-RNA sequencing to identify chemical perturbations that shift the transcriptional state of tumor necrosis factor (TNF)-activated TB-susceptible macrophages toward that of TB-resistant cells, i.e., prevents their aberrant activation without suppressing beneficial TNF responses. Last, we demonstrate that the compounds identified with this approach enhance the resistance of the TB-susceptible mouse macrophages to virulent Mycobacterium tuberculosis.

PMID:37756395 | DOI:10.1126/sciadv.adh4119

PLoS Genet. 2023 Sep 27;19(9):e1010937. doi: 10.1371/journal.pgen.1010937. eCollection 2023 Sep.

ABSTRACT

Telomeric repeat-containing RNA (TERRA) is a class of long noncoding RNAs (lncRNAs) that are transcribed from subtelomeric to telomeric region of chromosome ends. TERRA is prone to form R-loop structures at telomeres by invading into telomeric DNA. Excessive telomere R-loops result in telomere instability, so the TERRA level needs to be delicately modulated. However, the molecular mechanisms and factors controlling TERRA level are still largely unknown. In this study, we report that the RNA binding protein RBMX is a novel regulator of TERRA level and telomere integrity. The expression level of TERRA is significantly elevated in RBMX depleted cells, leading to enhanced telomere R-loop formation, replication stress, and telomere instability. We also found that RBMX binds to TERRA and the nuclear exosome targeting protein ZCCHC8 simultaneously, and that TERRA degradation slows down upon RBMX depletion, implying that RBMX promotes TERRA degradation by regulating its transportation to the nuclear exosome, which decays nuclear RNAs. Altogether, these findings uncover a new role of RBMX in TERRA expression regulation and telomere integrity maintenance, and raising RBMX as a potential target of cancer therapy.

PMID:37756323 | DOI:10.1371/journal.pgen.1010937

Toxics. 2023 Sep 18;11(9):788. doi: 10.3390/toxics11090788.

ABSTRACT

Aquatic environment are often contaminated with heavy metals from various industrial sources. However, physicochemical techniques for pollutant detection are limited, thus prompting the need for additional bioassays. We investigated the use of greater duckweed (Spirodela polyrhiza) as a bioindicator of metal pollution. We exposed S. polyrhiza to four pollutants (namely, silver, cadmium, copper, and chromium) and assessed metal toxicity by measuring its frond area and the length of its regrown roots. The plant displayed significant differences in both frond size and root growth in response to the four metals. Silver was the most toxic (EC50 = 23 µg L-1) while copper the least (EC50 = 365-607 µg L-1). Direct comparisons of metal sensitivity and the reliability of the two endpoint assays showed that root growth was more sensitive (lower in terms of 50% effective concentration) to chromium, cadmium, and copper, and was more reliable (lower in terms of coefficient of variation) than those for frond area. Compared to conventional Lemna-based tests, the S. polyrhiza test is easier to perform (requiring only one 24-well plate, 3 mL of medium and a 72-h exposure). Moreover, it does not require livestock cultivation/maintenance, making it more suitable for repeated measurements. Measurements of S. polyrhiza root length may be suitable for assessment when copper and chromium in municipal and industrial wastewater exceed the environmentally permissible levels.

PMID:37755798 | DOI:10.3390/toxics11090788

Proteomes. 2023 Aug 25;11(3):26. doi: 10.3390/proteomes11030026.

ABSTRACT

Individual gene expression and molecular pathway activation profiles were shown to be effective biomarkers in many cancers. Here, we used the human interactome model to algorithmically build 7470 molecular pathways centered around individual gene products. We assessed their associations with tumor type and survival in comparison with the previous generation of molecular pathway biomarkers (3022 "classical" pathways) and with the RNA transcripts or proteomic profiles of individual genes, for 8141 and 1117 samples, respectively. For all analytes in RNA and proteomic data, respectively, we found a total of 7441 and 7343 potential biomarker associations for gene-centric pathways, 3020 and 2950 for classical pathways, and 24,349 and 6742 for individual genes. Overall, the percentage of RNA biomarkers was statistically significantly higher for both types of pathways than for individual genes (p < 0.05). In turn, both types of pathways showed comparable performance. The percentage of cancer-type-specific biomarkers was comparable between proteomic and transcriptomic levels, but the proportion of survival biomarkers was dramatically lower for proteomic data. Thus, we conclude that pathway activation level is the advanced type of biomarker for RNA and proteomic data, and momentary algorithmic computer building of pathways is a new credible alternative to time-consuming hypothesis-driven manual pathway curation and reconstruction.

PMID:37755705 | DOI:10.3390/proteomes11030026

Curr Protoc. 2023 Sep;3(9):e883. doi: 10.1002/cpz1.883.

ABSTRACT

Calling Cards is a platform technology to record a cumulative history of transient protein-DNA interactions in the genome of genetically targeted cell types. The record of these interactions is recovered by next-generation sequencing. Compared with other genomic assays, readouts of which provide a snapshot at the time of harvest, Calling Cards enables correlation of historical molecular states to eventual outcomes or phenotypes. To achieve this, Calling Cards uses the piggyBac transposase to insert self-reporting transposon "Calling Cards" into the genome, leaving permanent marks at interaction sites. Calling Cards can be deployed in a variety of in vitro and in vivo biological systems to study gene regulatory networks involved in development, aging, and disease. Out of the box, it assesses enhancer usage but can be adapted to profile-specific transcription factor (TF) binding with custom TF-piggyBac fusion proteins. The Calling Cards workflow has five main stages: delivery of Calling Cards reagents, sample preparation, library preparation, sequencing, and data analysis. Here, we first present a comprehensive guide for experimental design, reagent selection, and optional customization of the platform to study additional TFs. Then, we provide an updated protocol for the five steps, using reagents that improve throughput and decrease costs, including an overview of a newly deployed computational pipeline. This protocol is designed for users with basic molecular biology experience to process samples into sequencing libraries in 2 days. Familiarity with bioinformatic analysis and command line tools is required to set up the pipeline in a high-performance computing environment and to conduct downstream analyses. © 2023 Wiley Periodicals LLC. Basic Protocol 1: Preparation and delivery of Calling Cards reagents Support Protocol 1: Next-generation sequencing quantification of barcode distribution within self-reporting transposon plasmid pool and adeno-associated virus genome Basic Protocol 2: Sample collection and RNA purification Support Protocol 2: Library density quantitative PCR Basic Protocol 3: Sequencing library preparation Basic Protocol 4: Library pooling and sequencing Basic Protocol 5: Data analysis.

PMID:37755132 | DOI:10.1002/cpz1.883

J Dev Biol. 2023 Aug 31;11(3):37. doi: 10.3390/jdb11030037.

ABSTRACT

Generating specialized cell types via cellular transcription factor (TF)-mediated reprogramming has gained high interest in regenerative medicine due to its therapeutic potential to repair tissues and organs damaged by diseases or trauma. Organ dysfunction or improper tissue functioning might be restored by producing functional cells via direct reprogramming, also known as transdifferentiation. Regeneration by converting the identity of available cells in vivo to the desired cell fate could be a strategy for future cell replacement therapies. However, the generation of specific cell types via reprogramming is often restricted due to cell fate-safeguarding mechanisms that limit or even block the reprogramming of the starting cell type. Nevertheless, efficient reprogramming to generate homogeneous cell populations with the required cell type's proper molecular and functional identity is critical. Incomplete reprogramming will lack therapeutic potential and can be detrimental as partially reprogrammed cells may acquire undesired properties and develop into tumors. Identifying and evaluating molecular barriers will improve reprogramming efficiency to reliably establish the target cell identity. In this review, we summarize how using the nematode C. elegans as an in vivo model organism identified molecular barriers of TF-mediated reprogramming. Notably, many identified molecular factors have a high degree of conservation and were subsequently shown to block TF-induced reprogramming of mammalian cells.

PMID:37754839 | DOI:10.3390/jdb11030037

mSphere. 2023 Sep 27:e0013423. doi: 10.1128/msphere.00134-23. Online ahead of print.

ABSTRACT

The gut microbiome is a potentially important mechanism that links prenatal disaster exposures with increased disease risks. However, whether prenatal disaster exposures are associated with alterations in the infant's gut microbiome remains unknown. We established a birth cohort study named Hurricane as the Origin of Later Alterations in Microbiome (HOLA) after Hurricane Maria struck Puerto Rico in 2017. We enrolled vaginally born Latino term infants aged 2 to 6 months, including n = 29 infants who were exposed in utero to Hurricane Maria in Puerto Rico and n = 34 infants who were conceived at least 5 months after the hurricane as controls. Shotgun metagenomic sequencing was performed on infant stool swabs. Infants exposed in utero to Hurricane Maria had a reduced diversity in their gut microbiome compared to the control infants, which was mainly seen in the exclusively formula-fed group (P = 0.02). Four bacterial species, including Bacteroides vulgatus, Clostridium innocuum, Bifidobacterium pseudocatenulatum, and Clostridium neonatale, were depleted in the exposure group compared to the control group. Compositional differences in the microbial community and metabolic genes between the exposure and control groups were significant, which were driven by the formula feeding group (P = 0.02 for the microbial community and P = 0.008 for the metabolic genes). Metabolic modules involved in carbohydrate metabolism were reduced in the exposure group. Prenatal maternal exposure to Hurricane Maria was associated with a reduced gut commensal and an altered microbial composition and metabolic potential in the offspring's gut. Breastfeeding can adjust the composition of the gut microbiomes of exposed infants. IMPORTANCE Climate change is a serious issue that is affecting human health. With more frequent and intense weather disasters due to climate change, there is an urgent need to evaluate and understand the impacts of prenatal disaster exposures on the offspring. The prenatal stage is a particularly vulnerable stage for disease origination. However, the impact of prenatal weather disaster exposures on the offspring's gut microbiome has not been evaluated. Our HOLA study starts to fill this knowledge gap and provides novel insights into the microbiome as a mechanism that links prenatal disaster exposures with elevated disease risks. Our major finding that reduced microbial diversity and altered metabolic capacity are associated with prenatal hurricane exposures warrants further studies to evaluate the impact of weather disasters on the unborn.

PMID:37754563 | DOI:10.1128/msphere.00134-23

Elife. 2023 Sep 27;12:e91362. doi: 10.7554/eLife.91362. Online ahead of print.

ABSTRACT

Drug resistance is a challenge in anticancer therapy. In many cases, cancers can be resistant to the drug prior to exposure, i.e., possess intrinsic drug resistance. However, we lack target-independent methods to anticipate resistance in cancer cell lines or characterize intrinsic drug resistance without a priori knowledge of its cause. We hypothesized that cell morphology could provide an unbiased readout of drug resistance. To test this hypothesis, we used HCT116 cells, a mismatch repair-deficient cancer cell line, to isolate clones that were resistant or sensitive to bortezomib, a well-characterized proteasome inhibitor and anticancer drug to which many cancer cells possess intrinsic resistance. We then expanded these clones and measured high-dimensional single-cell morphology profiles using Cell Painting, a high-content microscopy assay. Our imaging- and computation-based profiling pipeline identified morphological features that differed between resistant and sensitive cells. We used these features to generate a morphological signature of bortezomib resistance. We then employed this morphological signature to analyze a set of HCT116 clones (five resistant and five sensitive) that had not been included in the signature training dataset, and correctly predicted sensitivity to bortezomib in seven cases, in the absence of drug treatment. This signature predicted bortezomib resistance better than resistance to other drugs targeting the ubiquitin-proteasome system. Our results establish a proof-of-concept framework for the unbiased analysis of drug resistance using high-content microscopy of cancer cells, in the absence of drug treatment.

PMID:37753907 | DOI:10.7554/eLife.91362

PeerJ. 2023 Sep 22;11:e16129. doi: 10.7717/peerj.16129. eCollection 2023.

ABSTRACT

Metagenome binning is a key step, downstream of metagenome assembly, to group scaffolds by their genome of origin. Although accurate binning has been achieved on datasets containing multiple samples from the same community, the completeness of binning is often low in datasets with a small number of samples due to a lack of robust species co-abundance information. In this study, we exploited the chromatin conformation information obtained from Hi-C sequencing and developed a new reference-independent algorithm, Metagenome Binning with Abundance and Tetra-nucleotide frequencies-Long Range (metaBAT-LR), to improve the binning completeness of these datasets. This self-supervised algorithm builds a model from a set of high-quality genome bins to predict scaffold pairs that are likely to be derived from the same genome. Then, it applies these predictions to merge incomplete genome bins, as well as recruit unbinned scaffolds. We validated metaBAT-LR's ability to bin-merge and recruit scaffolds on both synthetic and real-world metagenome datasets of varying complexity. Benchmarking against similar software tools suggests that metaBAT-LR uncovers unique bins that were missed by all other methods. MetaBAT-LR is open-source and is available at https://bitbucket.org/project-metabat/metabat-lr.

PMID:37753177 | PMC:PMC10519199 | DOI:10.7717/peerj.16129

Biophys Rep. 2023 Apr 30;9(2):82-98. doi: 10.52601/bpr.2023.230007.

ABSTRACT

Phosphorylation is one of the most important post-translational modifications (PTMs) of proteins, governing critical protein functions. Most human proteins have been shown to undergo phosphorylation, and phosphoproteomic studies have been widely applied due to recent advancements in high-resolution mass spectrometry technology. Although the experimental workflow for phosphoproteomics has been well-established, it would be useful to optimize and summarize a detailed, feasible protocol that combines phosphoproteomics and data-independent acquisition (DIA), along with follow-up data analysis procedures due to the recent instrumental and bioinformatic advances in measuring and understanding tens of thousands of site-specific phosphorylation events in a single experiment. Here, we describe an optimized Phos-DIA protocol, from sample preparation to bioinformatic analysis, along with practical considerations and experimental configurations for each step. The protocol is designed to be robust and applicable for both small-scale phosphoproteomic analysis and large-scale quantification of hundreds of samples for studies in systems biology and systems medicine.

PMID:37753060 | PMC:PMC10518521 | DOI:10.52601/bpr.2023.230007

Biophys Rep. 2023 Apr 30;9(2):57-66. doi: 10.52601/bpr.2023.230005.

ABSTRACT

Identifying cancer-related differentially expressed genes provides significant information for diagnosing tumors, predicting prognoses, and effective treatments. Recently, deep learning methods have been used to perform gene differential expression analysis using microarray-based high-throughput gene profiling and have achieved good results. In this study, we proposed a new robust multiple-datasets-based semi-supervised learning model, MSSL, to perform tumor type classification and candidate cancer-specific biomarkers discovery across multiple tumor types and multiple datasets, which addressed the following long-lasting obstacles: (1) the data volume of the existing single dataset is not enough to fully exert the advantages of deep learning; (2) a large number of datasets from different research institutions cannot be effectively used due to inconsistent internal variances and low quality; (3) relatively uncommon cancers have limited effects on deep learning methods. In our article, we applied MSSL to The Cancer Genome Atlas (TCGA) and the Gene Expression Comprehensive Database (GEO) pan-cancer normalized-level3 RNA-seq data and got 97.6% final classification accuracy, which had a significant performance leap compared with previous approaches. Finally, we got the ranking of the importance of the corresponding genes for each cancer type based on classification results and validated that the top genes selected in this way were biologically meaningful for corresponding tumors and some of them had been used as biomarkers, which showed the efficacy of our method.

PMID:37753058 | PMC:PMC10518520 | DOI:10.52601/bpr.2023.230005

iScience. 2023 Sep 1;26(10):107810. doi: 10.1016/j.isci.2023.107810. eCollection 2023 Oct 20.

ABSTRACT

Research shows that brain circuits controlling vital physiological processes are closely linked with endogenous time-keeping systems. In this study, we aimed to examine oscillatory gene expression patterns of well-characterized neuronal circuits by reanalyzing publicly available transcriptomic data from a spatiotemporal gene expression atlas of a non-human primate. Unexpectedly, brain structures known for regulating circadian processes (e.g., hypothalamic nuclei) did not exhibit robust cycling expression. In contrast, basal ganglia nuclei, not typically associated with circadian physiology, displayed the most dynamic cycling behavior of its genes marked by sharp temporally defined expression peaks. Intriguingly, the mammillary bodies, considered hypothalamic nuclei, exhibited gene expression patterns resembling the basal ganglia, prompting reevaluation of their classification. Our results emphasize the potential for high throughput circadian gene expression analysis to deepen our understanding of the functional synchronization across brain structures that influence physiological processes and resulting complex behaviors.

PMID:37752952 | PMC:PMC10518731 | DOI:10.1016/j.isci.2023.107810

iScience. 2023 Aug 19;26(10):107678. doi: 10.1016/j.isci.2023.107678. eCollection 2023 Oct 20.

ABSTRACT

Soft tissue sarcomas (STSs) are rare malignant tumors that are difficult to prognosticate using currently available instruments. Omics sciences could provide more accurate and individualized survival predictions for patients with metastatic STS. In this pilot, hypothesis-generating study, we integrated clinicopathological variables with proton nuclear magnetic resonance (1H NMR) plasma metabolomic and lipoproteomic profiles, capturing both tumor and host characteristics, to identify novel prognostic biomarkers of 2-year survival. Forty-five metastatic STS (mSTS) patients with prevalent leiomyosarcoma and liposarcoma histotypes receiving trabectedin treatment were enrolled. A score combining acetate, triglycerides low-density lipoprotein (LDL)-2, and red blood cell count was developed, and it predicts 2-year survival with optimal results in the present cohort (84.4% sensitivity, 84.6% specificity). This score is statistically significant and independent of other prognostic factors such as age, sex, tumor grading, tumor histotype, frailty status, and therapy administered. A nomogram based on these 3 biomarkers has been developed to inform the clinical use of the present findings.

PMID:37752948 | PMC:PMC10518687 | DOI:10.1016/j.isci.2023.107678

iScience. 2023 Sep 9;26(10):107883. doi: 10.1016/j.isci.2023.107883. eCollection 2023 Oct 20.

ABSTRACT

The colonic mucosal barrier protects against infection, inflammation, and tissue ulceration. Composed primarily of Mucin-2, proteolytic erosion of this barrier is an invariant feature of colitis; however, the molecular mechanisms are not well understood. We have applied a recurrent food poisoning model of acquired inflammatory bowel disease using Salmonella enterica Typhimurium to investigate mucosal barrier erosion. Our findings reveal an innate Toll-like receptor 4-dependent mechanism activated by previous infection that induces Neu3 neuraminidase among colonic epithelial cells concurrent with increased Cathepsin-G protease secretion by Paneth cells. These anatomically separated host responses merge with the desialylation of nascent colonic Mucin-2 by Neu3 rendering the mucosal barrier susceptible to increased proteolytic breakdown by Cathepsin-G. Depletion of Cathepsin-G or Neu3 function using pharmacological inhibitors or genetic-null alleles protected against Mucin-2 proteolysis and barrier erosion and reduced the frequency and severity of colitis, revealing approaches to preserve and potentially restore the mucosal barrier.

PMID:37752945 | PMC:PMC10518488 | DOI:10.1016/j.isci.2023.107883

Cell Rep Methods. 2023 Sep 20:100597. doi: 10.1016/j.crmeth.2023.100597. Online ahead of print.

ABSTRACT

Decades of research have not yet fully explained the mechanisms of epithelial self-organization and 3D packing. Single-cell analysis of large 3D epithelial libraries is crucial for understanding the assembly and function of whole tissues. Combining 3D epithelial imaging with advanced deep-learning segmentation methods is essential for enabling this high-content analysis. We introduce CartoCell, a deep-learning-based pipeline that uses small datasets to generate accurate labels for hundreds of whole 3D epithelial cysts. Our method detects the realistic morphology of epithelial cells and their contacts in the 3D structure of the tissue. CartoCell enables the quantification of geometric and packing features at the cellular level. Our single-cell cartography approach then maps the distribution of these features on 2D plots and 3D surface maps, revealing cell morphology patterns in epithelial cysts. Additionally, we show that CartoCell can be adapted to other types of epithelial tissues.

PMID:37751739 | DOI:10.1016/j.crmeth.2023.100597

Am J Hum Genet. 2023 Sep 20:S0002-9297(23)00315-4. doi: 10.1016/j.ajhg.2023.09.002. Online ahead of print.

ABSTRACT

Congenital diaphragmatic hernia (CDH) is a relatively common and genetically heterogeneous structural birth defect associated with high mortality and morbidity. We describe eight unrelated families with an X-linked condition characterized by diaphragm defects, variable anterior body-wall anomalies, and/or facial dysmorphism. Using linkage analysis and exome or genome sequencing, we found that missense variants in plastin 3 (PLS3), a gene encoding an actin bundling protein, co-segregate with disease in all families. Loss-of-function variants in PLS3 have been previously associated with X-linked osteoporosis (MIM: 300910), so we used in silico protein modeling and a mouse model to address these seemingly disparate clinical phenotypes. The missense variants in individuals with CDH are located within the actin-binding domains of the protein but are not predicted to affect protein structure, whereas the variants in individuals with osteoporosis are predicted to result in loss of function. A mouse knockin model of a variant identified in one of the CDH-affected families, c.1497G>C (p.Trp499Cys), shows partial perinatal lethality and recapitulates the key findings of the human phenotype, including diaphragm and abdominal-wall defects. Both the mouse model and one adult human male with a CDH-associated PLS3 variant were observed to have increased rather than decreased bone mineral density. Together, these clinical and functional data in humans and mice reveal that specific missense variants affecting the actin-binding domains of PLS3 might have a gain-of-function effect and cause a Mendelian congenital disorder.

PMID:37751738 | DOI:10.1016/j.ajhg.2023.09.002

Cell Syst. 2023 Sep 22:S2405-4712(23)00244-2. doi: 10.1016/j.cels.2023.08.004. Online ahead of print.

ABSTRACT

Recent experimental developments in genome-wide RNA quantification hold considerable promise for systems biology. However, rigorously probing the biology of living cells requires a unified mathematical framework that accounts for single-molecule biological stochasticity in the context of technical variation associated with genomics assays. We review models for a variety of RNA transcription processes, as well as the encapsulation and library construction steps of microfluidics-based single-cell RNA sequencing, and present a framework to integrate these phenomena by the manipulation of generating functions. Finally, we use simulated scenarios and biological data to illustrate the implications and applications of the approach.

PMID:37751736 | DOI:10.1016/j.cels.2023.08.004

Cell Rep Methods. 2023 Sep 25;3(9):100576. doi: 10.1016/j.crmeth.2023.100576. Epub 2023 Aug 29.

ABSTRACT

The mammalian gut microbiome protects the host through colonization resistance (CR) against the incursion of exogenous and often harmful microorganisms, but identifying the exact microbes responsible for the gut microbiota-mediated CR against a particular pathogen remains a challenge. To address this limitation, we developed a computational method: generalized microbe-phenotype triangulation (GMPT). We first systematically validated GMPT using a classical population dynamics model in community ecology and demonstrated its superiority over baseline methods. We then tested GMPT on simulated data generated from the ecological network inferred from a real community (GnotoComplex microflora) and real microbiome data on two mouse studies on Clostridioides difficile infection. We demonstrated GMPT's ability to streamline the discovery of microbes that are potentially responsible for microbiota-mediated CR against pathogens. GMPT holds promise to advance our understanding of CR mechanisms and facilitate the rational design of microbiome-based therapies for preventing and treating enteric infections.

PMID:37751698 | DOI:10.1016/j.crmeth.2023.100576

Cell Rep Methods. 2023 Sep 25;3(9):100567. doi: 10.1016/j.crmeth.2023.100567. Epub 2023 Aug 28.

ABSTRACT

Chronological age prediction from DNA methylation sheds light on human aging, health, and lifespan. Current clocks are mostly based on linear models and rely upon hundreds of sites across the genome. Here, we present GP-age, an epigenetic non-linear cohort-based clock for blood, based upon 11,910 methylomes. Using 30 CpG sites alone, GP-age outperforms state-of-the-art models, with a median accuracy of ∼2 years on held-out blood samples, for both array and sequencing-based data. We show that aging-related changes occur at multiple neighboring CpGs, with implications for using fragment-level analysis of sequencing data in aging research. By training three independent clocks, we show enrichment of donors with consistent deviation between predicted and actual age, suggesting individual rates of biological aging. Overall, we provide a compact yet accurate alternative to array-based clocks for blood, with applications in longitudinal aging research, forensic profiling, and monitoring epigenetic processes in transplantation medicine and cancer.

PMID:37751697 | DOI:10.1016/j.crmeth.2023.100567

Cell Rep Methods. 2023 Sep 25;3(9):100577. doi: 10.1016/j.crmeth.2023.100577. Epub 2023 Aug 31.

ABSTRACT

The rapid accumulation of single-cell RNA-seq data has provided rich resources to characterize various human cell populations. However, achieving accurate cell-type annotation using public references presents challenges due to inconsistent annotations, batch effects, and rare cell types. Here, we introduce SELINA (single-cell identity navigator), an integrative and automatic cell-type annotation framework based on a pre-curated reference atlas spanning various tissues. SELINA employs a multiple-adversarial domain adaptation network to remove batch effects within the reference dataset. Additionally, it enhances the annotation of less frequent cell types by synthetic minority oversampling and fits query data with the reference data using an autoencoder. SELINA culminates in the creation of a comprehensive and uniform reference atlas, encompassing 1.7 million cells covering 230 distinct human cell types. We substantiate its robustness and superiority across a multitude of human tissues. Notably, SELINA could accurately annotate cells within diverse disease contexts. SELINA provides a complete solution for human single-cell RNA-seq data annotation with both python and R packages.

PMID:37751689 | DOI:10.1016/j.crmeth.2023.100577