Machine-Learning-Based m5C Score for the Prognosis Diagnosis of Osteosarcoma

RNA 5-Methylcytosine regulators are associated with cell adhesion and predict prognosis of endometrial cancer

Background

RNA methylation is a significant form of post-transcriptional modification that has been implicated in various diseases, including cancers. One prominent type of RNA methylation is 5-Methylcytosine (m5C), which primarily regulates RNA stability, transcription, and translation. However, the role of m5C-related gene regulation in cell adhesion within uterine corpus endometrial carcinoma (UCEC) remains unexplored. Therefore, the objective of this study was to investigate the association between RNA m5C methylation and UCEC and develop a prognostic predictive model to forecast survival outcomes in UCEC patients.

Methods

The RNA datasets were acquired from The Cancer Genome Atlas (TCGA) and Gene Expression Omnibus (GEO) databases. The dataset was used to explore the interaction relationships of m5C regulators in UCEC. Unsupervised clustering analysis identified clusters with distinct m5C modification patterns. Different clusters underwent Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment level analysis to investigate the effects of pathways related to m5C methylation, which were further validated through in vitro cellular experiments. A prognostic predictive model was developed using the least absolute shrinkage and selection operator (LASSO) and multivariate regression analysis.

Results

Two clusters with distinct m5C modification patterns were identified using unsupervised cluster analysis. Furthermore, the prognosis of cluster 2 was found to be worse. Enrichment analysis showed alterations in cell adhesion-related pathways in both clusters, as well as differences between the clusters. Through this analysis, we identified 25 genes with significant prognostic value. Finally, a prognostic predictive model comprising NSUN2 and YBX1 was constructed.

Conclusions

In conclusion, diverse m5C modification patterns display distinct cell adhesion properties in UCEC, which are correlated with prognosis and offer significant potential as prognostic markers for UCEC assessment. We developed a prognostic predictive model to accurately predict the prognosis of UCEC.

Identification of a novel MYC target gene set signature for predicting the prognosis of osteosarcoma patients

Osteosarcoma is a primary malignant tumor found mainly in teenagers and young adults. Patients have very little long-term survival. MYC controls tumor initiation and progression by regulating the expression of its target genes; thus, constructing a risk signature of osteosarcoma MYC target gene set will benefit the evaluation of both treatment and prognosis. In this paper, we used GEO data to download the ChIP-seq data of MYC to obtain the MYC target gene. Then, a risk signature consisting of 10 MYC target genes was developed using Cox regression analysis. The signature indicates that patients in the high-risk group performed poorly. After that, we verified it in the GSE21257 dataset. In addition, the difference in tumor immune function among the low- and high-risk populations was compared by single sample gene enrichment analysis. Immunotherapy and prediction of response to the anticancer drug have shown that the risk signature of the MYC target gene set was positively correlated with immune checkpoint response and drug sensitivity. Functional analysis has demonstrated that these genes are enriched in malignant tumors. Finally, STX10 was selected for functional experimentation. STX10 silence has limited osteosarcoma cell migration, invasion, and proliferation. Therefore, these findings indicated that the MYC target gene set risk signature could be used as a potential therapeutic target and prognostic indicator in patients with osteosarcoma.

Prognosis and Novel Drug Targets for Key lncRNAs of Epigenetic Modification in Colorectal Cancer

Background

Colorectal cancer (CRC) has been the 3rd most commonly malignant tumor of the gastrointestinal tract in the world. 5-Methylcytosine (m⁵C) and long noncoding RNAs (lncRNAs) have an essential role in predicting the prognosis and immune response for CRC patients. Therefore, we built a m⁵C-related lncRNA (m⁵CRlncRNA) model to investigate the prognosis and treatment methods for CRC patients.

Methods

Firstly, we secured the transcriptome and clinical data for CRC from The Cancer Genome Atlas (TCGA). Then, m⁵CRlncRNAs were recognized by coexpression analysis. Then, univariate Cox, least absolute shrinkage and selection operator (LASSO), and multivariate Cox regression analyses were utilized to build m⁵C-related prognostic characteristics. Besides, Kaplan-Meier analysis, ROC, PCA, C-index, enrichment analysis, and nomogram were performed to investigate the model. Additionally, immunotherapy responses and antitumor medicines were explored for CRC patients.

Results

A total of 8 m⁵C-related lncRNAs (AC093157.1, LINC00513, AC025171.4, AC090948.2, ZEB1-AS1, AC109449.1, AC009041.3, and LINC02516) were adopted to construct a risk model to investigate survival and prognosis for CRC patients. CRC samples were separated into low- and high-risk groups, with the latter having a worse prognosis. The m⁵C-related lncRNA model helps us to better distinguish immunotherapy responses and IC50 of antitumor medicines in different groups of CRC patients.

Conclusion

The research may give new perspectives on tailored therapy approaches as well as novel theories for forecasting the prognosis of CRC patients.

Comprehensive analysis of key m5C modification-related genes in type 2 diabetes

Background: 5-methylcytosine (m5C) RNA methylation plays a significant role in several human diseases. However, the functional role of m5C in type 2 diabetes (T2D) remains unclear.

Methods: The merged gene expression profiles from two Gene Expression Omnibus (GEO) datasets were used to identify m5C-related genes and T2D-related differentially expressed genes (DEGs). Least-absolute shrinkage and selection operator (LASSO) regression analysis was performed to identify optimal predictors of T2D. After LASSO regression, we constructed a diagnostic model and validated its accuracy. Gene ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) analyses were conducted to confirm the biological functions of DEGs. Gene Set Enrichment Analysis (GSEA) was used to determine the functional enrichment of molecular subtypes. Weighted gene co-expression network analysis (WGCNA) was used to select the module that correlated with the most pyroptosis-related genes. Protein-protein interaction (PPI) network was established using the STRING database, and hub genes were identified using Cytoscape software. The competitive endogenous RNA (ceRNA) interaction network of the hub genes was obtained. The CIBERSORT algorithm was applied to analyze the interactions between hub gene expression and immune infiltration.

Results: m5C-related genes were significantly differentially expressed in T2D and correlated with most T2D-related DEGs. LASSO regression showed that ZBTB4 could be a predictive gene for T2D. GO, KEGG, and GSEA indicated that the enriched modules and pathways were closely related to metabolism-related biological processes and cell death. The top five genes were identified as hub genes in the PPI network. In addition, a ceRNA interaction network of hub genes was obtained. Moreover, the expression levels of the hub genes were significantly correlated with the abundance of various immune cells.

Conclusion: Our findings may provide insights into the molecular mechanisms underlying T2D based on its pathophysiology and suggest potential biomarkers and therapeutic targets for T2D.