Decoding the lactylation-immune link in calcific aortic valve disease: a comprehensive bioinformatics approach

Figures

• 

  Fig. 1. Analysis of three GEO datasets after correction. A, B PCA plots showing the clustering of samples from three GEO datasets before (A) and after (B) batch correction; (C-D) Box plots illustrating the distribution of genes before (C) and after (D) normalization; (E) Volcano plot depicting the analysis of variance, where grey points represent genes with an adjusted p-value > 0.05; (F) Heatmap displaying the top 20 genes that differ between CAVD and Normal conditions

• 

  Fig. 2. Identification and functional analysis of lactation-related genes. A, B Venn diagrams showing the overlap between upregulated (A) and downregulated (B) genes with lactylation-related genes; (C) GO enrichment analysis of lactation-related genes, highlighting associated biological processes; D KEGG pathway enrichment analysis of lactation-related genes, focusing on relevant signaling pathways

• 

  Fig. 3. Identification of lactylation hub genes. A Lasso regression plot showing the optimal lambda value for model selection; B Importance ranking of selected features based on Random Forest analysis (top15); C Venn diagram illustrating the overlap of selected features from Lasso and Random Forest; D Chord diagram showing the correlation between selected features; E ROC curves of the eight genes for distinguishing CAVD from non-CAVD samples in the training set, with AUC values indicated. F ROC curves of the eight genes evaluated in an independent validation set, demonstrating their classification performance and generalizability

• 

  Fig. 4. Correlation between hub genes and KEGG pathways in CAVD. A–H GSEA analysis of lactylation-related hub genes (ARGUL1, PFN1, ALDH1A1, CRABP2, MBD2, ALB, SOD1, THRAP3) with respect to KEGG pathways

• 

  Fig. 5. Immune infiltration and hub gene analysis in CAVD. A Correlation analysis between different immune cell; B Box plot comparing the difference in immune infiltration levels between CAVD and control; C–J Correlation analysis between hub genes (PFN1, MBD2, CRABP2, ALDH1A1, TIIRAP3, ARGUL1, ALB, SOD1) and immune cell types

• 

  Fig. 6. Comprehensive analysis of single-cell data. A Clustering of cell types at different resolutions; B t-SNE plot showing cell type distribution; C Distribution of cell counts across clusters; D Differential gene expression analysis across clusters, highlighting significant genes (adjusted p-value < 0.01)

• 

  Fig. 7. Expression profiling of lactation-related genes based on scRNA-seq. A Expression and percent of expression of lactylation-related genes across different cell types; B t-SNE clustering based on cell type, lactylation score, and lactylation group; C Lactylation levels across various cell types; D Cell proportions for high and low lactylation groups; E–H Lactylation levels in mild and severe CAVD, including overall lactylation levels in mild and severe CAVD (E), and lactylation levels in VIC1 (F), M1 macrophages (G), and DC (H) in mild and severe CAVD

• 

  Fig. 8. Lactylation and hallmark gene set score correlation across different cell types in single-cell data. For each cell, a lactylation score and a hallmark gene set score were assigned, and their correlation was analyzed across distinct cell types. The heatmap shows the correlation between lactylation scores and hallmark gene set scores within each cell type, with color intensity representing the strength and direction (positive or negative) of the correlation

• 

  Fig. 9. Validation of hub gene expression and histopathological changes in the CAVD mouse model. A–H Relative expression levels of the 8 lactylation-related hub genes. I Histological sections of aortic valves from ApoE−/− mice fed a normal diet (ND) or a high-fat diet (HFD), including HE, Masson’s trichrome, and Von Kossa staining. J Pan-Kla immunofluorescence staining of aortic valve sections from ND and HFD groups. Data were presented as median and range. Statistical significance was determined using the Mann–Whitney U test. For graphical annotation, *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001