A novel five-gene metabolism-related risk signature for predicting prognosis and immune infiltration in endometrial cancer: A TCGA data mining

Performance comparison between multi-level gene expression data in cancer subgroup classification

Cancer is a serious disease that can affect various parts of the body such as breast, colon, lung or stomach. Each of these cancers has their own treatment dependent historical subgroups. Hence, the correct identification of cancer subgroup has almost same importance as the timely diagnosis of cancer. This is still a challenging task and a system with highest accuracy is essential. Current researches are moving towards analyzing the gene expression data of cancer patients for various purposes including biomarker identification and studying differently expressed genes, using gene expression data measured in a single level (selected from different gene levels including genome, transcriptome or translation). However, previous studies showed that information carried by one level of gene expression is not similar to another level. This shows the importance of integrating multi-level omics data in these studies. Hence, this study uses tumor gene expression data measured from various levels of gene along with the integration of those data in the subgroup classification of nine different cancers. This is a comprehensive analysis where four different gene expression data such as transcriptome, miRNA, methylation and proteome are used in this subgrouping and the performances between models are compared to reveal the best model.

Risk prediction model of uterine corpus endometrial carcinoma based on immune-related genes

BACKGROUND

Given the significant role of immune-related genes in uterine corpus endometrial carcinoma (UCEC) and the long-term outcomes of patients, our objective was to develop a prognostic risk prediction model using immune-related genes to improve the accuracy of UCEC prognosis prediction.

METHODS

The Limma, ESTIMATE, and CIBERSORT methods were used for cluster analysis, immune score calculation, and estimation of immune cell proportions. Univariate and multivariate analyses were utilized to develop a prognostic risk model for UCEC. Risk model scores and nomograms were used to evaluate the models. String constructs a protein-protein interaction (PPI) network of genes. The qRT-PCR, immunofluorescence, and immunohistochemistry (IHC) all confirmed the genes.

RESULTS

Cluster analysis divided the immune-related genes into four subtypes. 33 immune-related genes were used to independently predict the prognosis of UCEC and construct the prognosis model and risk score. The analysis of the survival nomogram indicated that the model has excellent predictive ability and strong reliability for predicting the survival of patients with UCEC. The protein-protein interaction network analysis of key genes indicates that four genes play a pivotal role in interactions: GZMK, IL7, GIMAP, and UBD. The quantitative real-time polymerase chain reaction (qRT-PCR), immunofluorescence, and immunohistochemistry (IHC) all confirmed the expression of the aforementioned genes and their correlation with immune cell levels. This further revealed that GZMK, IL7, GIMAP, and UBD could potentially serve as biomarkers associated with immune levels in endometrial cancer.

CONCLUSION

The study identified genes related to immune response in UCEC, including GZMK, IL7, GIMAP, and UBD, which may serve as new biomarkers and therapeutic targets for evaluating immune levels in the future.

A novel TCGA-validated programmed cell-death-related signature of ovarian cancer

BACKGROUND

Ovarian cancer (OC) is a gynecological malignancy tumor with high recurrence and mortality rates. Programmed cell death (PCD) is an essential regulator in cancer metabolism, whose functions are still unknown in OC. Therefore, it is vital to determine the prognostic value and therapy response of PCD-related genes in OC.

METHODS

By mining The Cancer Genome Atlas (TCGA), Genotype-Tissue Expression (GTEx) and Genecards databases, we constructed a prognostic PCD-related genes model and performed Kaplan-Meier (K-M) analysis and Receiver Operating Characteristic (ROC) curve for its predictive ability. A nomogram was created via Cox regression. We validated our model in train and test sets. Quantitative real-time PCR (qRT-PCR) was applied to identify the expression of our model genes. Finally, we analyzed functional analysis, immune infiltration, genomic mutation, tumor mutational burden (TMB) and drug sensitivity of patients in low- and high-risk group based on median scores.

RESULTS

A ten-PCD-related gene signature including protein phosphatase 1 regulatory subunit 15 A (PPP1R15A), 8-oxoguanine-DNA glycosylase (OGG1), HECT and RLD domain containing E3 ubiquitin protein ligase family member 1 (HERC1), Caspase-2.(CASP2), Caspase activity and apoptosis inhibitor 1(CAAP1), RB transcriptional corepressor 1(RB1), Z-DNA binding protein 1 (ZBP1), CD3-epsilon (CD3E), Clathrin heavy chain like 1(CLTCL1), and CCAAT/enhancer-binding protein beta (CEBPB) was constructed. Risk score performed well with good area under curve (AUC) (AUC3 - year =0.728, AUC5 - year = 0.730). The nomogram based on risk score has good performance in predicting the prognosis of OC patients (AUC1 - year =0.781, AUC3 - year =0.759, AUC5 - year = 0.670). Kyoto encyclopedia of genes and genomes (KEGG) analysis showed that the erythroblastic leukemia viral oncogene homolog (ERBB) signaling pathway and focal adhesion were enriched in the high-risk group. Meanwhile, patients with high-risk scores had worse OS. In addition, patients with low-risk scores had higher immune-infiltrating cells and enhanced expression of checkpoints, programmed cell death 1 ligand 1 (PD-L1), indoleamine 2,3-dioxygenase 1 (IDO-1) and lymphocyte activation gene-3 (LAG3), and were more sensitive to A.443,654, GDC.0449, paclitaxel, gefitinib and cisplatin. Finally, qRT-PCR confirmed RB1, CAAP1, ZBP1, CEBPB and CLTCL1 over-expressed, while PPP1R15A, OGG1, CASP2, CD3E and HERC1 under-expressed in OC cell lines.

CONCLUSION

Our model could precisely predict the prognosis, immune status and drug sensitivity of OC patients.

ECPC-IDS: A benchmark endometrial cancer PET/CT image dataset for evaluation of semantic segmentation and detection of hypermetabolic regions

BACKGROUND

Endometrial cancer is one of the most common tumors in the female reproductive system and is the third most common gynecological malignancy that causes death after ovarian and cervical cancer. Early diagnosis can significantly improve the 5-year survival rate of patients. With the development of artificial intelligence, computer-assisted diagnosis plays an increasingly important role in improving the accuracy and objectivity of diagnosis and reducing the workload of doctors. However, the absence of publicly available image datasets restricts the application of computer-assisted diagnostic techniques.

METHODS

In this paper, a publicly available Endometrial Cancer PET/CT Image Dataset for Evaluation of Semantic Segmentation and Detection of Hypermetabolic Regions (ECPC-IDS) are published. Specifically, the segmentation section includes PET and CT images, with 7159 images in multiple formats totally. In order to prove the effectiveness of segmentation on ECPC-IDS, six deep learning semantic segmentation methods are selected to test the image segmentation task. The object detection section also includes PET and CT images, with 3579 images and XML files with annotation information totally. Eight deep learning methods are selected for experiments on the detection task.

RESULTS

This study is conduct using deep learning-based semantic segmentation and object detection methods to demonstrate the distinguishability on ECPC-IDS. From a separate perspective, the minimum and maximum values of Dice on PET images are 0.546 and 0.743, respectively. The minimum and maximum values of Dice on CT images are 0.012 and 0.510, respectively. The target detection section's maximum mAP values on PET and CT images are 0.993 and 0.986, respectively.

CONCLUSION

As far as we know, this is the first publicly available dataset of endometrial cancer with a large number of multi-modality images. ECPC-IDS can assist researchers in exploring new algorithms to enhance computer-assisted diagnosis, benefiting both clinical doctors and patients. ECPC-IDS is also freely published for non-commercial at: https://figshare.com/articles/dataset/ECPC-IDS/23808258.

Potential biomarkers: The hypomethylation of cg18949415 and cg22193385 sites in colon adenocarcinoma

Overall cancer hypomethylation had been identified in the past, but it is not clear exactly which hypomethylation site is the more important for the occurrence of cancer. To identify key hypomethylation sites, we studied the effect of hypomethylation in twelve regions on gene expression in colon adenocarcinoma (COAD). The key DNA methylation sites of cg18949415, cg22193385 and important genes of C6orf223, KRT7 were found by constructing a prognostic model, survival analysis and random combination prediction a series of in-depth systematic calculations and analyses, and the results were validated by GEO database, immune microenvironment, drug and functional enrichment analysis. Based on the expression values of C6orf223, KRT7 genes and the DNA methylation values of cg18949415, cg22193385 sites, the least diversity increment algorithm were used to predict COAD and normal sample. The 100 % reliability and 97.12 % correctness of predicting tumor samples were obtained in jackknife test. Moreover, we found that C6orf223 gene, cg18949415 site play a more important role than KRT7 gene, cg22193385 site in COAD. In addition, we investigate the impact of key methylation sites on three-dimensional chromatin structure. Our results will be help for experimental studies and may be an epigenetic biomarker for COAD.