Lactylation associated biomarkers and immune infiltration in aortic dissection

Figures

• 

  Fig. 1. Data Preprocessing and Differential Analysis. (A-B) Normalization of expression profiles in the GSE153434 dataset. (C) Volcano plot of DEGs in the GSE153434 dataset, with upregulated genes shown in red and downregulated genes shown in green. (D) Heatmap of the top 20 most upregulated and the top 20 most downregulated DEGs in the GSE153434 dataset.

• 

  Fig. 2. Functional Enrichment and PPI Analysis. (A) Biological Process (BP) categories in the GO enrichment analysis of DEGs. (B) Cellular Component (CC) categories. (C) Molecular Function (MF) categories. (D) KEGG pathway enrichment analysis of DEGs. (E) Venn diagram showing the overlap between DEGs from the GSE153434 dataset and lactylation-related genes. (F) PPI network of lactylation-related DEGs.

• 

  Fig. 3. WGCNA Analysis and Key Module Selection.(A-B) Determination of the soft threshold for WGCNA analysis.(C) Identification of co-expression network modules.(D) Heatmap showing the correlation between WGCNA modules and AD.(E) Scatter plot of gene-module correlations with AD in WGCNA.

• 

  Fig. 4. Machine Learning and Optimal Feature Genes Screening.(A-B): LASSO regression identifies biomarkers for AD. (A): The trajectory of gene coefficients as the regularization parameter (log λ) increases, where less important gene coefficients shrink toward zero. (B): The 10-fold cross-validation curve, with the optimal λ value (λ_min) selected based on the minimum binomial deviance to achieve the best model performance with minimal overfitting. (C): RF identifies biomarkers for AD. (D): Venn diagram of biomarkers for AD identified by both machine learning and WGCNA. (E): ROC analysis showing the diagnostic efficiency of HMGA1 for AD. (F): ROC analysis showing the diagnostic efficiency of PGK1 for AD.

• 

  Fig. 5. Cross-validation of the optimal feature genes in an external dataset. (A-B): Expression levels of HMGA1 and PGK1 in the GSE52093 dataset. (C-D): Diagnostic efficiency of HMGA1 and PGK1 in the GSE52093 dataset.

• 

  Fig. 6. Immune infiltration and correlation analysis. (A) Stacked bar plot comparing 22 immune cell subsets between AD and NC samples. (B) Bar plot showing the differences in the proportions of 22 immune cell types between AD and NC samples. (C) Correlation analysis plots of HMGA1, PGK1, and immune infiltration (positive correlation coefficients indicate positive correlations, while negative coefficients indicate negative correlations).

• 

  Fig. 7. Single-cell sequencing analysis. (A): UMAP plot showing cell clustering. (B): UMAP plot depicting HMGA1 expression in each cluster. (C): UMAP plot depicting PGK1 expression in each cluster. (D): Violin plot showing HMGA1 expression across different cell groups. (E): Violin plot showing PGK1 expression across different cell groups.

• 

  Fig. 8. Validation of Optimal Feature Gene Expression in Patient Aortic Tissues. (A-C): Western blotting analysis of HMGA1 and PGK1 expression in aortic tissues from AD and NC groups, along with corresponding results. (D-G): Immunofluorescence analysis of HMGA1 and PGK1 expression in aortic tissues from AD and NC groups, along with corresponding results.

• 

  Fig. 9. Validation of the expression of optimal feature genes in animal experiments. (A) Gross appearance of the aorta in the control group and AD group mice. (B) Histopathological observations of the aorta in the control group and AD group mice. (C-D) Expression levels and statistical analysis of PGK1 in the aorta of the control group and AD group mice. (E-F) Expression levels and statistical analysis of HMGA1 in the aorta of the control group and AD group mice.