Exploring Hypertrophic Cardiomyopathy Biomarkers through Integrated Bioinformatics Analysis: Uncovering Novel Diagnostic Candidates

Figures

• 

  Figure 1. Workflow for identifying biomarkers in HCM.

• 

  Figure 2. Differential analysis results. (a) Volcano plot of differential genes in GSE36961. (b) Volcano plot of differential genes in GSE180313.

• 

  Figure 3. WGCNA analysis result. (a) Soft threshold analysis of GSE36961. Each power corresponds to scale independence and average connectivity. We selected an R cutoff of 0.89 and a soft threshold of 13 (red horizontal line) to construct a scale-free network topology. (b) Module-trait relationships in GSE36961. The numbers in the colored grids represent the correlation and their p values. Red indicates positive correlation, and blue indicates negative correlation. (c) Cluster dendrogram of GSE36961. Each branch represents one gene, and each color at the bottom represents a co-expression module. (d) Scatter plot of module membership and gene significance in the green module of GSE36961. (e) Scatter plot of module membership and gene significance in the grey module of GSE36961. (f) Soft threshold analysis of GSE180313. Each power corresponds to scale independence and average connectivity. We selected an R cutoff of 0.8 and a soft threshold of 5 (red horizontal line) to construct a scale-free network topology. (g) Module-trait relationships in GSE180313. The numbers in the colored grids represent the correlation and their p values. Red indicates positive correlation, and blue indicates…

• 

  Figure 4. (a) GO BP enrichment diagram of the module intersection genes in WGCNA. (b) GO CC enrichment diagram of the module intersection genes in WGCNA. (c) GO MF enrichment diagram of the module intersection genes in WGCNA. (d) KEGG enrichment diagram of the module intersection genes in WGCNA. (e) Intersection of genes from the green and grey modules in GSE36961 and genes from the turquoise module in GSE180313. (f) UMAP and TSNE diagram of GSE181764. (g) UMAP diagram of the top 6 highly variable genes in GSE181764. (h) Heatmap of the top 10 highly variable genes in 4 cell subtypes of GSE181764 distribution.

• 

  Figure 5. GO and KEGG enrichment diagrams of cardiomyocytes and fibroblasts in GSE181764. (a) KEGG enrichment diagram of cardiomyocytes. (b) GO MF enrichment diagram of cardiomyocytes. (c) GO CC enrichment diagram of cardiomyocytes. (d) GO BP enrichment diagram of cardiomyocytes. (e) KEGG enrichment diagram of fibroblasts. (f) GO MF enrichment diagram of fibroblasts. (g) GO CC enrichment diagram of fibroblasts. (h) GO BP enrichment diagram of fibroblasts.

• 

  Figure 6. Enrichment maps of natural killer T cells and endothelial cells in GSE181764. (a) KEGG enrichment map of natural killer T cells. (b) GO MF enrichment map of natural killer T cells. (c) GO CC enrichment map of natural killer T cells. (d) GO BP enrichment map of natural killer T cells. (e) KEGG enrichment map of endothelial cells. (f) GO MF enrichment map of endothelial cells. (g) GO CC enrichment map of endothelial cells. (h) GO BP enrichment map of endothelial cells.

• 

  Figure 7. (a) Cross-validation of the 7 predicted genes in the fitted model. (b) Optimal selection in the LASSO model. The dashed line on the left represents lambda.min, and the dashed line on the right represents lambda.1se. (c) Pearson analysis of the 7 predicted genes. (d) Differential expression of RTN4 and COL4A1 in the validation dataset GSE68316. (e) Differential expression of IER3 in the validation dataset GSE32453.

• 

  Figure 8. Highly effective translator (GNB) and support vector machine (SVM) algorithms predict whether RTN4, IER3, and COL4A1 play a key role in HCM. (a) AUC curve for the training set. (b) AUC curve for the test set.