Two lncRNA signatures with cuproptosis as a novel prognostic model and clinicopathological value for endometrioid endometrial adenocarcinoma.

OBJECTIVE: Cuproptosis may contribute to tumorigenesis. However, the predictive value and therapeutic significance of cuproptosis-related lncRNAs (CRLs) in endometrioid endometrial adenocarcinoma (EEA) remains unknown. METHODS: We obtained RNA-seq data from TCGA database and searched the Literature to identify cuproptosis-related genes. Using machine learning models, we identified prognostic lncRNAs for cuproptosis. Immune properties and drug sensitivity were investigated based on these signatures. Further, a ceRNA network was constructed by bioinformatics and in vitro experiments were performed. RESULTS: We determined two cuproptosis-related signatures to build the prognostic model in EEA. Afterward, the risk scores of two cuproptosis-related signatures were associated with clinicopathological molecular typing and as independent prognostic factors for EEA. In addition, we observed significant differences in immune function, checkpoints, and CD8+ T lymphocyte infiltration between the two risk groups. Furthermore, chemotherapy drugs such as AKT inhibitors exhibited lower IC50 values in the high-risk group. We speculate that ACOXL-AS1 can be served as an endogenous 'sponge' to regulate the expression of MTF1 by miR-421. Through in vitro experiments, we preliminarily validated the ceRNA network relationship in the cellular model. CONCLUSION: In EEAs, this study proposed a broad molecular signature of CRLs are promising biomarkers for predicting clinical outcomes and therapeutic responses.

Automated cervical cell segmentation using deep ensemble learning.

BACKGROUND: Cervical cell segmentation is a fundamental step in automated cervical cancer cytology screening. The aim of this study was to develop and evaluate a deep ensemble model for cervical cell segmentation including both cytoplasm and nucleus segmentation. METHODS: The Cx22 dataset was used to develop the automated cervical cell segmentation algorithm. The U-Net, U-Net + + , DeepLabV3, DeepLabV3Plus, Transunet, and Segformer were used as candidate model architectures, and each of the first four architectures adopted two different encoders choosing from resnet34, resnet50 and denseNet121. Models were trained under two settings: trained from scratch, encoders initialized from ImageNet pre-trained models and then all layers were fine-tuned. For every segmentation task, four models were chosen as base models, and Unweighted average was adopted as the model ensemble method. RESULTS: U-Net and U-Net + + with resnet34 and denseNet121 encoders trained using transfer learning consistently performed better than other models, so they were chosen as base models. The ensemble model obtained the Dice similarity coefficient, sensitivity, specificity of 0.9535 (95% CI:0.9534-0.9536), 0.9621 (0.9619-0.9622),0.9835 (0.9834-0.9836) and 0.7863 (0.7851-0.7876), 0.9581 (0.9573-0.959), 0.9961 (0.9961-0.9962) on cytoplasm segmentation and nucleus segmentation, respectively. The Dice, sensitivity, specificity of baseline models for cytoplasm segmentation and nucleus segmentation were 0.948, 0.954, 0.9823 and 0.750, 0.713, 0.9988, respectively. Except for the specificity of cytoplasm segmentation, all metrics outperformed the best baseline models (P < 0.05) with a moderate margin. CONCLUSIONS: The proposed algorithm achieved better performances on cervical cell segmentation than baseline models. It can be potentially used in automated cervical cancer cytology screening system.

Identification and validation of transferrin receptor protein 1 for predicting prognosis and immune infiltration in lower grade glioma.

Introduction: Transferrin receptor protein 1 (TFRC), an ananda molecule associated with ferroptosis, has been identified as affecting a wide spectrum of pathological processes in various cancers, but the prognostic value correlates with the tumor microenvironment of TFRC in lower-grade glioma (LGG) is still unclear. Materials and methods: Clinical pathological information and gene expression data of patients with LGG come from The Cancer Genome Atlas (TCGA), Chinese Glioma Genome Atlas (CGGA), GTEx, Oncomine, UCSC Xena, and GEO databases. We then used various bioinformatics methods and mathematical models to analyze those data, aiming to investigate the clinical significance of TFRC in LGG and illustrate its association with tumor immunity. In addition, the molecular function and mechanisms of TFRC were revealed by gene ontology (GO), Kyoto Encyclopedia of Genes and Genomes (KEGG), and gene set enrichment analysis (GSEA). Immunohistochemical experiments and single-cell analysis have been performed. Results: TFRC expression was highly expressed in many tumors and showed a poor prognosis. Including gliomas, it was significantly associated with several poor clinical prognostic variables, tumor immune microenvironment, tumor mutational burden (TMB), m6a modification, and ferroptosis in LGG. TFRC as a key factor was further used to build a prediction nomogram. The C-index, calibration curve, and decision curve analysis showed the nomogram was clinically useful and calibration was accurate. At the same time, we also demonstrated that promoter hypomethylation of DNA upstream of TFRC could lead to high TFRC expression and poor overall survival. There is a significant correlation between TFRC and CD8 + T cell, macrophage cell infiltration, and several immune checkpoints, such as PD-L1(cd274), CTLA4, and PD1, suggesting a novel direction for future clinical application. Functional and molecular mechanism analysis showed an association of TFRC expression with immune-related pathways through GSEA, GO, and KEGG analysis. Finally, immunohistochemical experiments and single-cell analysis confirmed the expression of TFRC in glioma. Conclusion: TFRC may be a potential prognostic biomarker and an immunotherapeutic target for glioma.