Efficient pulmonary-targeted therapy and mechanistic insights of a novel glutathione dry powder inhaler in a model of radiation-induced lung injury.

Iron overload triggers pathological remodeling of corneal stroma through ferroptosis and senescence-associated secretory phenotype in keratoconus.

Drug hypersensitivity reactions (DHRs) associated with a specific human leukocyte antigen (HLA) allotype do not manifest uniformly in all individuals within an HLA model population. This scenario highlights the complexity of predicting drug hypersensitivity reaction risk without considering additional factors. Furthermore, conducting prospective clinical studies of drug hypersensitivity reactions in humans is unfeasible. Therefore, this study aimed to prospectively evaluate HLA-mediated drug hypersensitivity reactions by using a unique transgenic mouse line harbouring HLA-B*57:01 (B*57:01-Tg) and performing abacavir-induced hypersensitivity reactions. B*57:01-Tg mice or their littermates were orally administered either an abacavir-containing or a vehicle-only diet. Subsequently, analyses were performed on isolated CD8+ T-cells collected from the lymph nodes, ear biopsies, or plasma of these mice. RNA sequencing and proteomic data analyses performed on the isolated CD8+ T-cells demonstrated that the expression levels of several glycolytic enzymes, including hexokinase 2, were significantly up-regulated in abacavir-treated B*57:01-Tg mice. Moreover, the glycolytic rate and metabolite production significantly increased in these mice. Although treatment with the pyruvic acid carrier inhibitor UK5099 did not affect CD8+ T-cell activation, inhibiting glycolysis using 2-deoxy-D-glucose or diet-based calorie restriction attenuated the increase in IFN-γ-secreting effector memory CD8+ T-cells. Consequently, this attenuation prevented CD8+ T-cell dermal infiltration in abacavir-treated B*57:01-Tg mice. Overall, these results suggest the indispensable role of glycolytic CD8+ T-cell metabolism in HLA-mediated abacavir-induced hypersensitivity reactions, which may further determine susceptibility to HLA-mediated drug hypersensitivity reactions in humans.

Sepsis is a dysregulated host response to infections, leading to organ dysfunction and posing a critical threat to human health. Despite tremendous progress in understanding the pathophysiology of sepsis, early diagnosis and clinical treatment efficacy remain unsatisfactory. This study aimed to identify transcriptomic alterations in peripheral blood mononuclear cells as potential biomarkers of sepsis. Bulk RNA-seq was performed on peripheral blood mononuclear cells obtained from 20 patients with sepsis and 12 healthy individuals. Multiple bioinformatics tools were used to identify key genes and signaling pathways associated with sepsis progression. The hub genes were further externally validated by publicly available blood transcriptomic data and experimentally verified by immunocytofluorescence assay. Differential expression analysis revealed 4,522 differentially expressed genes in patients with sepsis (n = 20) compared with healthy individuals (n = 12). Weighted gene coexpression network analysis identified multiple gene modules closely related to sepsis, with the royal blue module exhibiting the most positive correlation with sepsis. Intersection analysis yielded 176 common genes between the royal blue module genes and differentially expressed genes. Protein-protein interaction analysis revealed five hub genes ( CTSB , CTSD , ATP6V0D1 , UBE2D1 , and ATP6V0C ) associated with sepsis. Immune infiltration was dissected by single-sample gene set enrichment analysis, revealing associations between hub genes and monocytes. Single-cell RNA sequencing data analysis and immunocytofluorescence assay confirmed the upregulation of CTSB and ATP6V0D1 in circulating monocytes. Notably, CTSB and ATP6V0D1 were significantly associated with 28-day mortality of sepsis patients in the external validation cohort (n = 479). This study identifies CTSB and ATP6V0D1 expression in circulating monocytes as potential biomarkers and promising therapeutic targets for sepsis.

Sulfhydrated TFEB alleviates blast-induced lung injury by maintaining epithelial barrier integrity through the IGF2R/MMP-2/9 pathway.

Single-cell sequencing has revolutionized biomedical research by offering insights into cellular heterogeneity at unprecedented resolution. Yet, the low signal-to-noise ratio characteristic of single-cell RNA sequencing (scRNA-seq) challenges quantitative analyses. Gene regulatory network (GRN) analysis can help overcome this obstacle, enabling the mechanistic elucidation of cellular state determinants. For instance, the VIPER algorithm can identify Master Regulator proteins from gene expression data. However, as the size and complexity of scRNA-seq datasets grow, the demand for scalable tools supporting the analysis of datasets with up to hundreds of thousands of cells becomes increasingly critical in its original implementation in R. RESULTS: To address this challenge, we introduce pyVIPER, a Python-based tool for protein activity inference from transcriptional data. pyVIPER supports flexible data transformation/postprocessing modules, enrichment analysis algorithms, and features a novel data structure for GRNs manipulation. It integrates seamlessly with scverse, scanpy and widely adopted machine learning libraries. By leveraging PyTorch-based GPU acceleration and optimized core operations, benchmarking demonstrates orders-of-magnitude improvements in runtime efficiency compared to R-based VIPER, reducing analysis time for large datasets from hours to minutes. CONCLUSIONS: pyVIPER is a fast, memory-efficient, and highly scalable Python toolkit for protein activity inference in large-scale scRNA-seq datasets. Its scalability and hardware acceleration enables high-throughput VIPER-based analysis of virtually any single-cell dataset while facilitating integration with other Python-based, including state-of-the-art machine learning workflows. Taken together, these features make pyVIPER a valuable resource to expand the applicability of mechanistic regulatory network-based analysis in single-cell research.

A central challenge in multi-condition single-cell RNA sequencing (scRNA-seq) data analysis is the disentanglement of true biological signals from unwanted variations in complex experimental designs. Current statistical and machine learning-based methods struggle with this task, often providing only visualizable embeddings, over-correcting and discarding biological signal, or failing to resolve cell-type-specific responses. Here, we present CAPER, a matrix factorization framework that explicitly disentangles shared biological states from condition-specific variations. CAPER directly outputs an interpretable, batch-corrected expression matrix in which the signal of interest is preserved and isolated. The performance of CAPER is validated using extensive simulations, followed by three real-world multi-condition scRNA-seq data applications, representing distinct signal-to-noise ratio (SNR) scenarios: a controlled immune stimulation in PBMCs with high SNR, a tumor-microenvironment dataset from LUAD with confounded SNR, and a complex autoimmune disease dataset from T1D with low SNR. Across these settings, CAPER yields interpretable latent factors linked to relevant biology, accurately recovers key differentially expressed genes, and correctly identifies the most responsive cell populations. CAPER is a robust and interpretable tool for recovering biological signals from multi-condition single-cell RNA-seq data, enabling reliable discovery in disease research and functional genomics.

Fibrosis resulting from metabolic-associated steatohepatitis (MASH) is increasingly recognized as the predominant form of liver fibrosis. Although the activation of hepatic stellate cells (HSCs) is essential for liver fibrosis, the mechanisms underlying HSC activation in MASH remain inadequately understood. Integrated analysis of large-scale single-cell and single-nucleus RNA sequencing data from human healthy and fibrosis samples reveals a distinct subpopulation of HSCs in MASH. AREL1 is a characteristic gene of this subpopulation and is uniquely upregulated in MASH-related fibrosis. HSC-specific knockout of Arel1 markedly attenuates liver fibrosis in MASH model male mice. Mechanistically, AREL1 is regulated by cholesterol and facilitates HSC activation through the AREL1-ILK axis, subsequently activating the PI3K-AKT signaling pathway. Moreover, therapeutic knockdown of Arel1 using vitamin A-modified lipid nanoparticles markedly ameliorates MASH-related liver fibrosis. Here, we show a unique mechanism underlying HSC activation in MASH-driven fibrosis and present the targeted knockdown of AREL1 in HSCs as a therapeutic avenue.

Human Idiopathic Pulmonary Fibrosis (IPF) is a progressive and fatal lung disease with unknown etiology and lacking efficient treatments. Here, we reported that human urine stem cells (hUSCs) significantly alleviated pulmonary fibrosis via inhibiting macrophage-myofibroblast transition (MMT), which was identified as a pivotal pathological process in IPF, with the strong interaction among infiltrated macrophages, damaged alveolar epithelial cells, and myofibroblasts via single-nucleus RNA sequencing data analysis and co-immunostaining. In addition, hUSCs significantly alleviated pulmonary fibrosis by attenuating alveolar epithelial cell damage, reducing monocyte-derived macrophage infiltration, and suppressing MMT in the bleomycin-induced pulmonary fibrosis mouse model. Furthermore, we demonstrated hUSCs inhibited monocyte recruitment and MMT via paracrine actions in the macrophage-alveolar epithelial cell co-culture system. Mechanistically, DKK1, which was highly secreted by hUSCs and identified by Venn diagram analysis between the luminex assay in supernatants of THP1 treated with hUSC-CM and antibody array of hUSC-CM, might contribute to preventing MMT via suppressing the Wnt/β-catenin signaling pathway in macrophages. In summary, hUSCs exerted multifaceted protective effects against pulmonary fibrosis, at least in part through paracrine mechanisms involving DKK1 and its modulation of Wnt/β-catenin-associated fibrotic responses in MMT. Therefore, hUSCs might provide a potential therapeutic strategy for IPF clinically.

Background:Alcohol-associated hepatitis (AH) is a severe inflammatory liver condition driven by dysregulated immune responses. CD8 T cells accumulate in the liver during AH, yet their functional role remains poorly defined.Methods:A murine NIAAA chronic-binge model with lipopolysaccharide (LPS) challenge was used. Single-cell RNA sequencing of liver and blood immune cells defined CD8 T cell subsets, followed by pathway analysis and reanalysis of human AH datasets. Spatial relationships were validated by immunohistochemistry of human AH liver tissue with CD8 TRM and liver sinusoidal endothelial cell (LSEC) co-staining. LSEC–CD8 T cell interactions were examined using in vitro co-culture, and signaling pathways were assessed by bulk RNA sequencing and phospho-protein analyses. IL15 function was evaluated by in vivo neutralization.Results:Single-cell RNA sequencing in the murine AH model revealed a distinct population of CD8 tissue-resident memory T cells (TRM) with heightened activation and proinflammatory cytokine production. Analysis of human AH liver single-cell RNA sequencing data, along with immunohistochemistry validation, supported CD8 TRM enrichment in diseased tissue. IL15 emerged as a prominent pathway linked to TRM activation, and IL15 blockade reduced TRM abundance and attenuated EtOH/ lipopolysaccharide-induced liver injury. Mechanistically, LSECs not only provided a structural niche for TRM retention but also amplified IL15 signalling. In vitro, coculture experiments demonstrated that LSECs intensified activation in pre-stimulated CD8 T cells in a stimulus-dependent manner: under IL15 stimulation, LSECs boosted effector function without inducing cell death, whereas under T-cell receptor stimulation, LSECs drove hyperactivation and activation-induced cell death. Bulk RNA-seq and phospho-protein analysis identified the PI3K-AKT pathway as a shared pathway enhanced by LSEC coculture in activated CD8 T cells. These findings define a context-dependent mechanism in which LSECs promote IL15-driven signaling through AKT pathway amplification, promoting TRM persistence and inflammatory activity in AH.Conclusions:IL15-associated signaling within the hepatic microenvironment, shaped by LSEC-CD8 T-cell interactions, promotes activation and persistence of CD8 TRM cells in AH. Targeting the IL15-LSEC-AKT axis may disrupt pathogenic TRM niches and represent a promising therapeutic strategy for severe AH.

Lung adenocarcinoma (LUAD) is one of the most severe malignant tumors. Phosphoinositides metabolism (PIM) plays an important role in maintaining the normal life activities of the organism and regulating tumor development. This study aimed to comprehensively investigate the association between LUAD and PIM. Data on LUAD and PIM-related genes (PIM-RGs) were sourced from public databases. Differential expression, univariate Cox regression analyses, and machine learning were conducted to identify prognostic genes. A risk model was subsequently developed, and LUAD patients were categorized into a high-risk group (HRG) and a low-risk group (LRG). Independent prognostic factors for LUAD were identified, and a nomogram was constructed. Functional enrichment, tumor microenvironment, mutations, and drug sensitivity analyses were also conducted to investigate the molecular mechanisms underlying LUAD. Additionally, single-cell RNA sequencing (scRNA-seq) data analysis was employed to identify key cells and clarify the dynamics of prognostic genes' expression. Ultimately, prognostic gene expression was investigated in clinical samples. MTMR7, GDPD1, MTMR4, and MTMR10 were recognized as prognostic genes. The risk model and nomogram (incorporating risk score and Stage) had good predictive performance. Notably, LUAD's malignant progression might be closely associated with biological processes including cellular protein synthesis, abnormal activation of neuroactive ligand-receptor interactions, and anti-tumor immune responses. Additionally, there was a general positive correlation between the differentially infiltrated immune cells of HRG and LRG. Moreover, TP53 and TTN had relatively high mutation frequencies in both HRG and LRG, and 142 drugs exhibited differential sensitivity between the 2 groups. Interestingly, epithelial cells were identified as LUAD's key cell type, with prognostic gene expression showing dynamic changes as these cells differentiated. Consistently, compared with the control group, GDPD1 and MTMR4 were upregulated, MTMR10 was downregulated, and MTMR7 showed no statistical difference but a certain upward trend in the LUAD group. This study identified 4 prognostic genes and constructed an effective risk model, providing a new perspective on the treatment of LUAD.

ABSTRACTRegulatory evolution can alter phenotypes, but cis- and trans-regulatory mechanisms may also diverge extensively while total transcript abundance remains stable. Comparisons of parental expression with allele-specific expression in F1 hybrids provide a framework for separating cis- and trans-regulatory effects because both parental alleles are measured in a shared trans-regulatory environment. Here, we analyzed RNA sequencing data fromSaccharomyces cerevisiae,Saccharomyces paradoxus, and their F1 hybrid. Regulatory divergence was widespread, with 61.3% of tested orthologs showing significant divergence in at least one cis or trans component. However, hybrid expression remained largely conserved, with 81.6% of genes not significantly different from either parent. Compensatory cis-trans divergence predominated over reinforcing divergence, consistent with widespread buffering of transcript abundance. To connect genome-wide patterns to mechanism, we analyzed the strongly cis-diverged locusLYS2and found species differences in promoter architecture, including anS. cerevisiae-specific AT-rich insertion, altered spacing among candidate regulatory features, and a promoter-proximal TATA-like element unique toS. cerevisiae. Sequence-based nucleosome prediction suggests that these differences create a broader promoter-proximal nucleosome-depleted region inS. cerevisiaethan inS. paradoxus. We also quantified allele-resolved intron retention and found that splicing was broadly conserved, with only rare locus-specific hybrid-associated shifts. Together, these results show that regulatory divergence is widespread but often buffered in the hybrid, whereas post-transcriptional divergence is comparatively limited.

Most colorectal cancer (CRC) patients exhibit resistance to immune checkpoint blockade (ICB), limiting treatment efficacy. Activating the unfolded protein response sensor IRE1α in cancer cells can induce anticancer immune responses, yet its regulation remains unclear. Here we identify Dolichyl-Phosphate Mannosyltransferase 1 (DPM1) as a regulator of IRE1 expression and activity using BioID screen. Analysis of CRC patient RNA-sequencing data reveals that low DPM1 expression correlates with an IRE1-dependent transcriptional signature, increased immune infiltration, and improved ICB responses. Mechanistically, DPM1 ablation reduces protein glycosylation, causing chronic IRE1 activation in cancer cells and enhanced cytotoxic T cell-mediated immunosurveillance. Inhibition or knock-out of IRE1 reverses this effect. These findings establish DPM1 as a modulator of IRE1 activity that influences tumor immunogenicity, suggesting its potential as a therapeutic target to improve cancer immunotherapy outcomes.

Cell type identification is a fundamental step in the analysis of single-cell RNA sequencing (scRNA-seq) data. Among the various classification tools, support vector machine (SVM) classifiers have traditionally demonstrated strong overall performance. However, the rapid accumulation of scRNA-seq data has led to a significant increase in SVM training time. Convolutional neural networks (CNNs), known for their success in image recognition and their ability to handle high-dimensional data, offer a promising alternative. Nevertheless, the inherent translation invariance of CNNs proves counterproductive in the context of scRNA-seq data, often resulting in misclassification of cell types. To address this limitation, we propose a novel cell type identification method, termed BP-Coord, which incorporates coordinate information as additional channels to enhance the spatial awareness of the model. Furthermore, a bicubic interpolation upsampling layer is introduced prior to the CoordConv layers, enabling the CNN to capture more precise positional information and better adapt to translation variations in the data. Experimental results on five public scRNA-seq benchmark datasets demonstrate that the proposed BP-Coord model consistently outperforms state-of-the-art methods, including SVM-based classifiers and recent deep learning approaches such as SuperCT and scGAC. In particular, BP-Coord achieves accuracy improvements of up to 3.5 % over the best competing methods on large-scale PBMC datasets and shows superior robustness on imbalanced and small-sample datasets. These results highlight the effectiveness of incorporating explicit positional encoding into convolutional architectures for automatic cell type identification.

Differential expression (DE) analysis of RNA sequencing (RNA-Seq) data are cornerstone of transcriptomic research. Widely used statistical frameworks are primarily optimized to detect monotonic mean shifts between conditions and may therefore overlook genes or microRNAs whose disease association arises at both low and high expression levels. Such non-monotonic patterns, referred to here as improper expression profiles, may reflect biologically relevant heterogeneity but remain difficult to identify using standard tools. To evaluated whether receiver operating characteristic (ROC)-based indices, specifically the generalized area under the curve (gAUC) and the length of the ROC curve (LROC), can support exploratory screening and prioritization of improper expression profiles in RNA-Seq data, as a complement to conventional DE methods. Methodological study. Using simulated negative binomial count data, we compared DESeq2, classical AUC (cAUC), gAUC, and LROC across varying sample sizes and dispersion levels, focusing on improper expression profiles. Performance was summarized using true positive rate and positive predictive value under ranking-based feature selection, including a one-shot benchmark operating point (available only in simulations) and sensitivity analyses across selection sizes. The methods were also applied to a publicly available CC miRNA dataset using heuristic post-hoc screening rules informed by simulation diagnostics. cAUC was largely insensitive to improper expression patterns. DESeq2 performed robustly for conventionally differentially expressed features but recovered a smaller fraction of simulated improper profiles under ranking-based selection. Across simulation scenarios, gAUC showed the highest and most stable recovery of improper profiles, whereas LROC provided complementary signal under low-to-moderate dispersion but degraded under extreme overdispersion. In the CC dataset, ROC-derived indices identified candidate improper miRNAs that were not prioritized by DESeq2, and several top candidates had literature support consistent with biological plausibility. gAUC, supported by LROC as an auxiliary index, provides a practical ROC-based screening extension to standard RNA-Seq workflows. Because these indices are applied using heuristic thresholds without controlled error rates, the resulting candidates should be interpreted as exploratory prioritization and require independent validation.

Severe acute pancreatitis (SAP) is often associated with life-threatening acute lung injury (ALI), with its pathogenesis being intricately connected to dysregulated inflammatory responses. There is a deficiency of particular treatment options. Dexmedetomidine (DEX), a highly selective α2-adrenergic receptor (α2-AR) agonist, demonstrates not only sedative effects but also possesses anti-inflammatory and organ-protective properties. However, the mechanisms through which DEX exerts its effects in SAP and related pulmonary disorders remain uncertain. A rat SAP model was established via retrograde injection of 5% sodium taurocholate into the biliopancreatic ducts, with DEX intervention and positive control groups included. The effects of DEX on pathological damage to pancreatic and lung tissues, serum inflammatory factors, and pulmonary edema were evaluated in vivo. In vitro, lipopolysaccharide (LPS) stimulation of human umbilical vein endothelial cells (HUVECs) was used to model an inflammatory environment. Transcriptome sequencing, single-cell RNA sequencing data analysis, protein-protein interaction network construction, molecular docking, and molecular biology techniques were employed to investigate the action targets of DEX and its regulatory effects on the absent in melanoma 2 (AIM2) inflammasome signaling pathway. DEX therapy attenuated the pathological injury to pancreatic and pulmonary tissues in SAP rats, decreased serum concentrations of amylase, interleukin-1 beta, and tumor necrosis factor-alpha, and suppressed pulmonary edema. Transcriptomic analysis revealed that DEX could partially reverse the disorder in lung tissue gene expression profiles induced by SAP and identified the "DEX target gene set". Bioinformatics analysis identified AIM2 as the core target, and molecular docking indicated that DEX could bind to AIM2 effectively. Single-cell RNA sequencing analysis revealed that the target gene set was specifically highly expressed in endothelial cells. DEX inhibited the activation of AIM2 and Caspase-11, as well as the phosphorylation of the nuclear factor kappa-B signaling pathway in lung tissue and endothelial cells. It also decreased the expression of vascular cell adhesion protein-1 and matrix metalloproteinase-9 in lung tissue. In addition, AIM2 knocdown blocked LPS-induced apoptosis in HUVECs, and DEX showed no further inhibitory effect. This study identifies a novel mechanism in which DEX may alleviate pyroptosis of endothelial cells by targeting and inhibiting AIM2 inflammasome activation, thereby improving SAP-ALI.

Clear cell adenocarcinoma of the urinary tract (CCA-UT) is a rare, potentially aggressive tumor with very limited information regarding its clinicopathologic characteristics and molecular alterations. This study aimed to elucidate the clinicopathologic features and molecular landscape of one of the largest cohorts (35 cases) of this tumor, to identify genomic alterations and potential therapeutic targets. Seventy-nine percent of the patients were women, with a median age of 61 years. The urethra was the most common site (18; 51%), and all cases were ≥pT2 (pT2:15; pT3:11; pT4:8). Twenty-nine percent of the patients died of their disease on follow-up. On whole-exome sequencing, pathogenic/oncogenic alterations were identified in 91% (32/35) cases. These alterations, most frequently involved chromatin modifiers (66% cases), including ATRX , KMT2C , ARID1A , and ARID1B . Other frequently mutated genes included ATM , NF1 , and ERBB2 . Ninety-seven percent (34/35) of cases were microsatellite stable, and tumor mutational burden (TMB) was >10 mut/Mb in 9% (3/35) of cases. Five cases were homologous recombinant-deficient on ScarHRD analysis, and 3 cases showed BRCA mutations. Recurrent copy number loss events in Chr 1(p36.33-p35.3) were the most common copy number alterations (80%; n=28 cases). RNA-sequencing data analysis revealed numerous differentially expressed genes and enrichment of the epithelial-to-mesenchymal transition gene signature in individual samples. However, there was no statistical significance in the progression-free survival between cases with epithelial and mesenchymal phenotypes. CCA-UT are aggressive tumors with a heterogeneous molecular profile, underscoring the role of molecular analysis in identifying potential therapeutic options for the treatment of this pernicious tumor.

Colorectal cancer (CRC) is a major health concern and understanding its molecular characteristics is crucial for improving its diagnosis and treatment. Here, we present a comprehensive analysis utilizing RNA-sequencing (RNA-seq) data and clinical information from Korean patients with CRC. Differential gene expression analysis identified significant changes in gene expression between tumor and normal tissues. Gene Set Enrichment Analysis (GSEA) revealed dysregulated pathways associated with tumor progression. Furthermore, using CMScaller, we successfully stratified CRC tissues into distinct molecular subtypes. Upon reviewing the public consensus molecular subtype (CMS) signature, it was confirmed that it shares similar biological characteristics with the existing CRC. Additionally, biological characteristics of the group that could not be classified using CMScaller were found to resemble those of CMS2. Finally, distinguishing characteristics were observed between the tumor and normal groups when analyzed from an immunological perspective. Patients with CRC were checked for immunotherapy responsiveness, and those who clinically responded to immunotherapy were identified. Survival analysis confirmed that certain microsatellite stable (MSS) samples were responsive to immunotherapy and showed a relatively better prognosis. Furthermore, analysis of various immune cell types to identify genes involved in the response to immunotherapy revealed that RORC, NOS2, and KLRK1 are potential candidate genes. Our findings provide valuable insights into the molecular landscape of CRC in the Korean population and underscore the potential for integrating RNA-seq data with clinical information to improve cancer research and patient care. Immunotherapy was found to be effective in Korean patients with CRC.

Chimeric RNAs resulting from intergenic splicing represent a distinct mechanism for transcriptome expansion. To explore the role of this previously unidentified layer of the transcriptome in sex-specific immunity, we analyzed RNA sequencing data from 425 blood samples and identified a female-specific chimeric RNA, UBA1-CDK16, which was further validated in more than 1200 blood samples. This chimeric RNA forms via cis-splicing between two adjacent X-linked parental genes, UBA1 and CDK16, despite both being expressed in both sexes. We demonstrated that a female-specific chromatin loop at the UBA1-CDK16 junction sites facilitates the intergenic splicing. Evolutionary analysis revealed that UBA1-CDK16 became female specific in humans through at least two independent paths. Functional studies suggested that UBA1-CDK16 is enriched in the myeloid lineage and may regulate myeloid cell development. Notably, its abnormal expression in female patients with COVID-19 correlates with altered neutrophil counts, highlighting its potential role in the disease progression.

Identifying different cell types is a prerequisite step in the analysis of single-cell RNA sequencing (scRNA-seq) data, with clustering being a common technique utilized for this purpose. However, high dropout rates inherent in scRNA-seq data and complex intercellular relationships become main challenges in scRNA-seq data analysis. To address these issues, we proposed a novel model based on zero-inflated negative binomial (ZINB) distribution and graph attention network for scRNA-seq data clustering (scZGA). scZGA consists of three key modules. The first module captures the global probabilistic structure using a ZINB model. The second module constructs the graph with Pearson's correlation coefficient, and employs a graph autoencoder with residual connection to learn important neighbor relationships while preserving topological structure information simultaneously. The final module conducts deep clustering through a self-optimizing embedding algorithm. With these improvements, clustering results show that scZGA consistently achieves higher scores across six scRNA-seq datasets by using evaluation metrics such as normalized mutual information and adjusted rand index.