scRNA-seq revealed high stemness epithelial malignant cell clusters and prognostic models of lung adenocarcinoma.

Lung adenocarcinoma (LUAD) is one of the sole causes of death in lung cancer patients. This study combined with single-cell RNA-seq analysis to identify tumor stem-related prognostic models to predict the prognosis of lung adenocarcinoma, chemotherapy agents, and immunotherapy efficacy. mRNA expression-based stemness index (mRNAsi) was determined by One Class Linear Regression (OCLR). Differentially expressed genes (DEGs) were detected by limma package. Single-cell RNA-seq analysis in GSE123902 dataset was performed using Seurat package. Weighted Co-Expression Network Analysis (WGCNA) was built by rms package. Cell differentiation ability was determined by CytoTRACE. Cell communication analysis was performed by CellCall and CellChat package. Prognosis model was constructed by 10 machine learning and 101 combinations. Drug predictive analysis was conducted by pRRophetic package. Immune microenvironment landscape was determined by ESTIMATE, MCP-Counter, ssGSEA analysis. Tumor samples have higher mRNAsi, and the high mRNAsi group presents a worse prognosis. Turquoise module was highly correlated with mRNAsi in TCGA-LUAD dataset. scRNA analysis showed that 22 epithelial cell clusters were obtained, and higher CSCs malignant epithelial cells have more complex cellular communication with other cells and presented dedifferentiation phenomenon. Cellular senescence and Hippo signaling pathway are the major difference pathways between high- and low CSCs malignant epithelial cells. The pseudo-temporal analysis shows that cluster1, 2, high CSC epithelial cells, are concentrated at the end of the differentiation trajectory. Finally, 13 genes were obtained by intersecting genes in turquoise module, Top200 genes in hdWGCNA, DEGs in high- and low- mRNAsi group as well as DEGs in tumor samples vs. normal group. Among 101 prognostic models, average c-index (0.71) was highest in CoxBoost + RSF model. The high-risk group samples had immunosuppressive status, higher tumor malignancy and low benefit from immunotherapy. This work found that malignant tumors and malignant epithelial cells have high CSC characteristics, and identified a model that could predict the prognosis, immune microenvironment, and immunotherapy of LUAD, based on CSC-related genes. These results provided reference value for the clinical diagnosis and treatment of LUAD.

The role of T cells and shared genes in psoriasis and inflammatory bowel disease based on single-cell RNA and comprehensive analysis.

Psoriasis and inflammatory bowel disease (IBD) have a similar etiology, including abnormal activation of T cells. Differentially expressed genes (DEGs) analysis was used to search for shared genes. GO (Gene Ontology) and KEGG (Kyoto Encyclopedia of Genes and Genomes) analysis were then performed. Secondly, single-cell RNA analysis (scRNA-seq) and immune infiltration were employed to explore the immune imbalance of the diseases. By weighted gene co expression network analysis (WGCNA), we obtained hub shared genes. Furthermore, we analyzed the diagnostic performance and immune association with the hub genes. Finally, functional enrichment of miRNAs related to hub shared genes was carried out. Single-cell analysis showed a high proportion of T cells among infiltrated immune cells and immune infiltration showed CD4+ T and γδ T cells were significantly elevated in diseases. Hub shared genes, LCN2, CXCL1 and PI3 had excellent diagnostic properties and were positively correlated with neutrophils, CD4+ T and γδ T cells. IL17 and TNF signaling pathway were the common pathway. In conclusion, CD4+ and γδ T cells and hub shared genes may play a crucial part in common mechanism between psoriasis and IBD. Moreover, hub shared genes may be potential diagnostic markers.

Structure-Aware Long Short-Term Memory Network for 3D Cephalometric Landmark Detection.

Detecting 3D landmarks on cone-beam computed tomography (CBCT) is crucial to assessing and quantifying the anatomical abnormalities in 3D cephalometric analysis. However, the current methods are time-consuming and suffer from large biases in landmark localization, leading to unreliable diagnosis results. In this work, we propose a novel Structure-Aware Long Short-Term Memory framework (SA-LSTM) for efficient and accurate 3D landmark detection. To reduce the computational burden, SA-LSTM is designed in two stages. It first locates the coarse landmarks via heatmap regression on a down-sampled CBCT volume and then progressively refines landmarks by attentive offset regression using multi-resolution cropped patches. To boost accuracy, SA-LSTM captures global-local dependence among the cropping patches via self-attention. Specifically, a novel graph attention module implicitly encodes the landmark's global structure to rationalize the predicted position. Moreover, a novel attention-gated module recursively filters irrelevant local features and maintains high-confident local predictions for aggregating the final result. Experiments conducted on an in-house dataset and a public dataset show that our method outperforms state-of-the-art methods, achieving 1.64 mm and 2.37 mm average errors, respectively. Furthermore, our method is very efficient, taking only 0.5 seconds for inferring the whole CBCT volume of resolution 768×768×576 .

CircUBXN7 suppresses cell proliferation and facilitates cell apoptosis in lipopolysaccharide-induced cell injury by sponging miR-622 and regulating the IL6ST/JAK1/STAT3 axis.

Acute respiratory distress syndrome (ARDS) is a common and serious respiratory illness with substantial morbidity and mortality. Circular RNAs have been demonstrated to participate in various diseases processes. However, the biological function and mechanism of most circular RNAs have not been elucidated in ARDS. In this study, we found that circUBXN7 was significantly increased in lipopolysaccharide (LPS)-induced A549 and Beas-2B cell injury. Inhibition of circUBXN7 significantly promoted cell proliferation and reduced cell apoptosis, while overexpression of circUBXN7 suppressed cell proliferation and accelerated cell apoptosis in LPS-induced A549 and Beas-2B cells. CircUBXN7 acted as a sponge for miR-622, and miR-622 rescued the effect of circUBXN7 on cell proliferation and apoptosis. We also found that IL6ST was a target gene of miR-622, and the expression of IL6ST was indirectly regulated by circUBXN7. Furthermore, western blotting indicated that the JAK1/STAT3 signaling pathway was involved in the circUBXN7/miR-622/IL6ST axis in LPS-induced A549 and Beas-2B cell injury. Overall, our study suggested that circUBXN7 suppressed cell proliferation and facilitated cell apoptosis by sponging miR-622 and regulating IL6ST, to activate the JAK1/STAT3 signaling pathway in LPS-induced A549 and Beas-2B cell injury. CircUBXN7 might therefore be a potential biomarker for ARDS, and dysregulation of circUBXN7 may be involved in the pathogenesis of ARDS.

TSegNet: An efficient and accurate tooth segmentation network on 3D dental model.

Automatic and accurate segmentation of dental models is a fundamental task in computer-aided dentistry. Previous methods can achieve satisfactory segmentation results on normal dental models; however, they fail to robustly handle challenging clinical cases such as dental models with missing, crowding, or misaligned teeth before orthodontic treatments. In this paper, we propose a novel end-to-end learning-based method, called TSegNet, for robust and efficient tooth segmentation on 3D scanned point cloud data of dental models. Our algorithm detects all the teeth using a distance-aware tooth centroid voting scheme in the first stage, which ensures the accurate localization of tooth objects even with irregular positions on abnormal dental models. Then, a confidence-aware cascade segmentation module in the second stage is designed to segment each individual tooth and resolve ambiguities caused by aforementioned challenging cases. We evaluated our method on a large-scale real-world dataset consisting of dental models scanned before or after orthodontic treatments. Extensive evaluations, ablation studies and comparisons demonstrate that our method can generate accurate tooth labels robustly in various challenging cases and significantly outperforms state-of-the-art approaches by 6.5% of Dice Coefficient, 3.0% of F1 score in term of accuracy, while achieving 20 times speedup of computational time.

Structure-Driven Unsupervised Domain Adaptation for Cross-Modality Cardiac Segmentation.

Performance degradation due to domain shift remains a major challenge in medical image analysis. Unsupervised domain adaptation that transfers knowledge learned from the source domain with ground truth labels to the target domain without any annotation is the mainstream solution to resolve this issue. In this paper, we present a novel unsupervised domain adaptation framework for cross-modality cardiac segmentation, by explicitly capturing a common cardiac structure embedded across different modalities to guide cardiac segmentation. In particular, we first extract a set of 3D landmarks, in a self-supervised manner, to represent the cardiac structure of different modalities. The high-level structure information is then combined with another complementary feature, the Canny edges, to produce accurate cardiac segmentation results both in the source and target domains. We extensively evaluate our method on the MICCAI 2017 MM-WHS dataset for cardiac segmentation. The evaluation, comparison and comprehensive ablation studies demonstrate that our approach achieves satisfactory segmentation results and outperforms state-of-the-art unsupervised domain adaptation methods by a significant margin.