Integrative Analysis of Mitochondrial‐Related Genes Reveals Diagnostic Biomarkers and Therapeutic Targets in Acute Pancreatitis

Figures

• 

  FIGURE 1. Flow chart of the study.

• 

  FIGURE 2. Identification of AP and mitochondrial related hub genes. (A) The volcano plot of differential expression analysis shows significant gene changes between the AP group and the normal group. (B) Venn diagram of the MRGs and the DEGs of AP. (C) Construct PPI for 34 intersecting genes based on STRING database. (D) Perform GeneMANIA analysis on intersecting genes. (E) Functional enrichment analysis of intersection genes based on Metascape database.

• 

  FIGURE 3. MRGs‐related subgroups in AP. (A) A consensus clustering matrix with k = 3. (B) The expression characteristics of the three subtypes of AP MRGs. (C) PCA with three subtypes of distribution. (D–E) Comparison of immune cell infiltration between three clusters. (Kruskal–Wallis test, ns, nonsignificant, *p < 0.05, **p < 0.01, and ***p < 0.001) (F) The Sankey diagram shows the matching of three subgroups with clinical severity classification.

• 

  FIGURE 4. Determine the model genes of AP using machine learning techniques. (A) Selecting 9 genes through LASSO regression. (B) Identification of 13 genes using SVM. (C–D) RF evaluated the importance of all genes to obtain 4 genes with importance scores greater than 3.

• 

  FIGURE 5. Diagnostic performance of the model on AP. (A) The Venn diagram of the intersection genes of three machine learning models yields three model genes. (B) The box plot shows the expression differences of MT‐ND5, MTX1 and NDUFA1 between AP and normal samples (Wilcoxon rank‐sum test, ***p < 0.001). (C–D) ROC curves of three genes and models. (E–F) The expression levels of inflammatory factors and three model genes in the cell model. (t test, ns, nonsignificant, *p < 0.05, **p < 0.01, and ***p < 0.001) (G–I) The expression levels of three genes in MAP and NMAP. (t test, ns, nonsignificant, *p < 0.05, and ****p < 0.0001).

• 

  FIGURE 6. Correlation analysis between model genes and immune cells. (A) The correlation between different types of immune cells and MT‐ND5. The correlation between MT‐ND5 and (B) neutrophils and (C) NK cells resting. (D) The correlation between different types of immune cells and MTX1. The correlation between MTX1 and (E) neutrophils, (F) T cells CD4 memory resting. (G) The correlation between different types of immune cells and NDUFA1. The correlation between NDUFA1 and (H) neutrophils, (I) T cells CD4 memory resting. (Spearman rank correlation, p < 0.05 considered statistically significant).

• 

  FIGURE 7. TFs, immunofluorescence, and single‐cell localisation analysis of model genes. (A–C) Predict TFs for MT‐ND5, MTX1 and NDUFA1. (D) Intersection Venn diagram of TFs of three model genes. (E–F) Immunofluorescence of MTX1 and NDUFA1 based on HPA database; (G–I) Use HPA database for single‐cell mapping of MT‐ND5, MTX1 and NDUFA1.

• 

  FIGURE 8. Molecular docking and MD simulation analysis of MTX1‐talniflumete and NDUFA1‐ABT‐737 complex. (A) The molecular docking structure of MTX1‐talniflumete, with blue dashed lines representing hydrogen bonds; (B) The molecular docking structure of NDUFA1‐ABT‐737, with hydrogen bonds as shown by the blue dashed line; (C) RMSD plot of MTX1 talniflumete; (D) RMSD diagram of NDUFA1‐ABT‐737; (E) RMSF plot of MTX1 talniflumete; (F) RMSF diagram of NDUFA1‐ABT‐737; (G) MTX1 talniflumete SASA; (H) SASA of NDUFA1‐ABT‐737; (I) The turning radius (Rg) of MTX1 talniflumete; (J) Rg of NDUFA1‐ABT‐737; number of hydrogen bonds of (K) MTX1 talniflumete; (L) NDUFA1‐ABT‐737 hydrogen bond quantity; (M) Two dimensional and three‐dimensional FEL of MTX1 talniflumete; (N) Two dimensional and three‐dimensional FEL of NDUFA1‐ABT‐737 system; (O) Principal Component Analysis (PCA) covariance matrix heatmap of MTX1 talniflumete; (Q) PCA covariance matrix heatmap of NDUFA1‐ABT‐737; (P) FEL based on PC1 and PC2 MTX1 talniflumete; (R) FEL of NDUFA1‐ABT‐737 based on PC1 and PC2.