Characterization of PANoptosis-based markers for ischemic stroke and their appealing implications in improving prognosis and in glioma patients

Fig. 1. Differentially expressed PANoptosis genes in peripheral blood of ischemic stroke. (A) Volcano diagram illustrating the DEGs in peripheral blood of ischemic stroke (IS) versus that of control subjects. (B) Venn chart of the shared DEGs and PANoptosis genes. (C) Circle map depicting the genomic position of the differentially expressed PANoptosis genes. (D) Heatmap of the transcriptional values of the differentially expressed PANoptosis genes in peripheral blood from ischemic stroke and controls. (E) Box plots of the differential expression of these genes in peripheral blood of ischemic stroke in comparison to that of controls. (F, G) GO and KEGG pathway enrichment results on the differentially expressed PANoptosis genes. (H) Heatmap illustrating the interactions between them. *p < 0.05; **p < 0.01; ***p < 0.001

Fig. 2. Classification of ischemic stroke into two PANoptosis molecule subtypes, and analysis their relevant molecular mechanisms and immune cell infiltration. (A) Consensus matrix when k = 2 in accordance with the peripheral blood expression of the differentially expressed PANoptosis genes. (B) PCA plots proving the distinction of the two subtypes at the transcriptional levels. (C) Heatmap of the expression of the differentially expressed PANoptosis genes in peripheral blood from the two subtypes. (D) Box plots illustrating the differential expression of these genes in the two subtypes. (E, F) Differential GO and KEGG terms between the C1 and C2 subtypes. (G) Heatmap illustrating the abundance of immune cell types in peripheral blood from ischemic stroke (IS) and control subjects. (H) Comparison of the infiltration levels of immune cells between ischemic stroke and control subjects. (I) Heatmap of the crosstalk between diverse immune cell populations. (J, K) The differential infiltration of diverse immune cell types in the two subtypes. *p < 0.05; **p < 0.01; ***p < 0.001

Fig. 3. Identification of PANoptosis molecule subtyping-related genes. (A) Dendrogram of ischemic stroke peripheral blood specimens (upper) and heatmap of the two PANoptosis molecule subtypes (lower). (B) Scale independence or mean connectivity under a gradient of soft threshold values. (C) Selection of the optimal threshold of soft connectivity. (D) Gene dendrogram and establishment of co-expression modules. (E) Heatmap illustrating the relationships of the merged modules with the PANoptosis molecule subtyping. (F) Comparison of gene significance across the modules. (G) Scatter diagram depicting the association of module membership of the yellow module with gene significance of the PANoptosis molecule subtyping. (H-J) Biological processes, cellular components, and molecule functions of the yellow module genes. (K) KEGG pathway enrichment results on the yellow module genes

Fig. 4. Multiple machine learning algorithms for selecting PANoptosis signature genes for ischemic stroke diagnosis. (A) Venn chart of the shared DEGs and the yellow module genes. (B) Box plots showing the root mean square of residuals in four machine learning algorithms: GLM, XGB, RF, and SVM. (C) Reverse cumulative distribution of |residuals| in the four machine learning algorithms. (D) The feature importance constructed by the four machine learning algorithms. (E) ROC curves of the four machine learning models in diagnosing ischemic stroke. (F) Box plots depicting the differential expression of the key genes in peripheral blood from ischemic stroke (IS) and control subjects. (G) Box plots for the differential expression of the key genes in the two PANoptosis molecule subtypes. ***p < 0.001

Fig. 5. Evaluation of each PANoptosis signature genes in diagnosing ischemic stroke, and construction of a diagnostic nomogram. (A-H) ROC curves of the PANoptosis signature genes: TMEM55A, CCPG1, CYP1B1, PJA2, MTPN, SCYL2, BAZ2B, and DPYD in diagnosing ischemic stroke. (I) The nomogram comprising PANoptosis signature genes for ischemic stroke risk prediction. (J, K) Calibration and decision curves for evaluating the accuracy of the nomogram in predicting ischemic stroke risk

Fig. 6. Signaling pathways underlying the PANoptosis signature genes. (A-H) GSEA for probing the signaling pathways related to BAZ2B, CCPG1, CYP1B1, DPYD, MTPN, PJA2, SCYL2, and TMEM55A

Fig. 7. Associations of the PANoptosis signature genes with immune cell infiltration in peripheral blood of ischemic stroke. (A-H) Correlation analyses on BAZ2B, CCPG1, CYP1B1, DPYD, MTPN, PJA2, SCYL2, and TMEM55A with the infiltration of immune cell populations

Fig. 8. Survival analysis on BAZ2B, CCPG1, CYP1B1, DPYD, MTPN, TMEM55A (PIP4P2), PJA2, and SCYL2 in glioma

Fig. 9. ROC curve analysis of the eight genes in glioma. (A) Prognostic ROC curve analysis was performed to evaluated the prognostic value of PANoptosis molecule subtyping-related genes as 1-year overall survival in glioma patients. (B) ROC analysis was performed to evaluated the diagnostic value of PANoptosis molecule subtyping-related genes between normal and glioma tissue

Fig. 10. The protein expression level of eight genes was evaluated in HPA database

Fig. 11. Immunofluorescence revealed the localization of MTPN and PJA2 in the U-251MG glioma cell