Bioinformatics-based screening and validation of PANoptosis-related biomarkers in periodontitis

Abstract

Background: Periodontitis is the most prevalent chronic inflammatory disease affecting the periodontal tissues. PANoptosis, a recently characterized form of programmed cell death, has been implicated in various pathological processes; however, its mechanistic role in periodontitis remains unclear. This study integrates multi-omics data and machine learning approaches to systematically identify and validate key PANoptosis-related biomarkers in periodontitis.

Methods: Periodontitis-related microarray datasets (GSE16134 and GSE10334) were obtained from the GEO database, and PANoptosis-related genes were retrieved from GeneCards. Differential gene expression analysis was performed using the GSE16134 dataset, followed by weighted gene co-expression network analysis (WGCNA) to identify relevant gene modules. The intersection of differentially expressed genes and WGCNA modules was used to define differentially expressed PANoptosis-related genes (PRGs). Protein-protein interaction (PPI) networks of these PRGs were constructed using the STRING database and visualized with Cytoscape. Subnetworks were identified using the MCODE plugin. Key genes were selected based on integration with rank-sum test results. Functional enrichment analysis was performed for these key genes. Machine learning algorithms were then applied to screen for potential biomarkers. Diagnostic performance was assessed using receiver operating characteristic (ROC) curves and box plots. The relationship between selected biomarkers and immune cell infiltration was explored using the CIBERSORT algorithm. Finally, RT-qPCR was conducted to validate biomarker expression in clinical gingival tissue samples.

Results: Through comprehensive bioinformatics analysis and literature review, ZBP1 was identified as a PANoptosis-related biomarker in periodontitis. RT-qPCR validation demonstrated that ZBP1 expression was significantly elevated in periodontitis tissues compared to healthy periodontal tissues (P < 0.05).

Conclusion: This study provides bioinformatic evidence linking PANoptosis to periodontitis. ZBP1 was identified as a key PANoptosis-related biomarker, suggesting that periodontitis may involve activation of the ZBP1-mediated PANoptosome complex.

Keywords: PANoptosis; ZBP1; bioinformatics; biomarkers; periodontitis.

FIGURE 1

Differential gene analysis results. (A) Volcano plot of DEGs; (B) Heatmap of DEGs.

FIGURE 2

GSVA analysis results. (A) Boxplot of PRGs-GSVA score differences between the periodontitis group and the healthy control group; (B) Boxplot of rank sum test for differential expression of PRGs (*P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001).

FIGURE 3

WGCNA analysis results. (A) Gene hierarchical clustering diagram; (B) Module correlation heatmap; (C) Venn diagram of WGCNA-PRGs and DEGs intersection.

FIGURE 4

(A) PPI network of Cluster 1; (B) PPI network of 48 key genes.

FIGURE 5

(A) GO enrichment analysis of key genes; (B) KEGG enrichment analysis of key genes.

FIGURE 6

Machine learning identification of potential biomarkers. (A) LASSO regression coefficients; (B) Cross-validation curve; (C) SVM result accuracy; (D) SVM result error rate; (E) Random forest ntree value selection; (F) Variable importance ranking; (G) Venn diagram of the intersection of results from the three machine learning methods.

FIGURE 7

Differential expression analysis and ROC evaluation of candidate biomarkers. (A,B) Training set results; (C,D) Validation set results.

FIGURE 8

Immune infiltration analysis results. (A) Relative percentage of immune cell subpopulations in samples; (B) Differences in immune cell infiltration in samples (*P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001); (C) Biomarker-immune cell correlations (training set); (D) Biomarker-immune cell correlations (validation set).

FIGURE 9

RT-qPCR results of ZBP1 (**P < 0.01).