Identification of a disulfidptosis-related genes signature for diagnostic and immune infiltration characteristics in cervical cancer

Abstract

Background: Cervical cancer (CC) ranks as the fourth most common malignancy affecting women globally, with research highlighting a rising incidence among younger age groups. Disulfidptosis, a newly identified form of regulated cell death, has been implicated in the pathogenesis of numerous diseases. This study employs bioinformatics analyses to explore the expression profiles and functional roles of disulfidptosis-related genes (DRGs) in the context of cervical cancer.

Methods: Differential analysis of the gene expression matrix in CC was performed to identify differentially expressed genes. The overlap between these genes and disulfidptosis-related genes was then determined. Key hub genes were identified using multiple machine learning approaches, including LASSO regression, support vector machines (SVM), and random forest (RF). These hub genes were subsequently used to construct a predictive model, which was validated using external datasets to ensure robustness and reliability.

Results: In this study, 11 overlapping genes were identified, among which four hub genes-BRK1, NDUFA11, RAC1, and NDUFS1-were extracted using machine learning techniques. The diagnostic performance of these hub genes was validated with external datasets, and a predictive model was constructed based on their expression. The model demonstrated an exceptionally high area under the curve (AUC) of 0.997. Moreover, AUC values exceeding 0.85 for two independent validation datasets further confirmed the model's accuracy and stability. Notably, NDUFA11 and BRK1 showed significant associations with patient survival, highlighting their prognostic importance in cervical squamous cell carcinoma. Using CMAP and DGIdb databases, Metformin and Coenzyme-I were identified as potential targeted therapies for NDUFS1 and NDUFA11, respectively, offering new therapeutic avenues for patients.

Conclusion: This study uncovered a strong association between disulfidptosis and CC and developed a predictive model to assess the risk in CC patients. These findings offer novel insights into identifying biomarkers and potential therapeutic targets for CC, paving the way for improved diagnostic and treatment strategies.

Fig 1. The study flow chart.

Fig 2. Differential expression analysis.

(A) A volcano plot of the DEGs. (B) KEGG pathway enrichment analysis of DEGs. (C) A Venn diagram illustrating the overlap between DEGs and DRGs. (D) A heatmap depicting the expression of 11 de-DRGs based on the GSE63514 dataset.

Fig 3. PPI network analysis (A) PPI network of 11 de-DRGs.

(B) Pearson correlation analysis of the 11 de-DRGs.

Fig 4. Immune infiltration of 11 de-DRGs.

(A) Boxplot of 11 de-DRGs expression levels. (B) The fraction of immune cells comparison in CC and normal group. (C) The correlation between de-DRGs and immune cells.

Fig 5. Screen of disulfidptosis-related signatures.

(A) LASSO regression. (B) SVM. (C) RF. (D) A Venn diagram illustrates the intersection of candidate genes identified by the three methods.

Fig 6. Validation of the diagnostic efficacy of hub genes.

(A) Correlation analysis of the four hub genes reveals their relationships, with positive correlations shown in red and negative correlations in blue. (B) A nomogram was constructed to predict the risk of CC based on the identified hub genes. (C) ROC curves were generated to evaluate the diagnostic performance of the hub genes in CC. (D) A calibration curve was used to assess the prediction accuracy of the nomogram model. (E-G) ROC curves from external validation datasets, including GSE63514, GSE67522, and GSE52903, respectively, further confirmed the diagnostic efficacy of the model.

Fig 7. The Kaplan-Meier survival analysis of four genes extracted from CESC.