Identification and validation of aging-related genes in heart failure based on multiple machine learning algorithms

Abstract

Background: In the face of continued growth in the elderly population, the need to understand and combat age-related cardiac decline becomes even more urgent, requiring us to uncover new pathological and cardioprotective pathways.

Methods: We obtained the aging-related genes of heart failure through WGCNA and CellAge database. We elucidated the biological functions and signaling pathways involved in heart failure and aging through GO and KEGG enrichment analysis. We used three machine learning algorithms: LASSO, RF and SVM-RFE to further screen the aging-related genes of heart failure, and fitted and verified them through a variety of machine learning algorithms. We searched for drugs to treat age-related heart failure through the DSigDB database. Finally, We use CIBERSORT to complete immune infiltration analysis of aging samples.

Results: We obtained 57 up-regulated and 195 down-regulated aging-related genes in heart failure through WGCNA and CellAge databases. GO and KEGG enrichment analysis showed that aging-related genes are mainly involved in mechanisms such as Cellular senescence and Cell cycle. We further screened aging-related genes through machine learning and obtained 14 key genes. We verified the results on the test set and 2 external validation sets using 15 machine learning algorithm models and 207 combinations, and the highest accuracy was 0.911. Through screening of the DSigDB database, we believe that rimonabant and lovastatin have the potential to delay aging and protect the heart. The results of immune infiltration analysis showed that there were significant differences between Macrophages M2 and T cells CD8 in aging myocardium.

Conclusion: We identified aging signature genes and potential therapeutic drugs for heart failure through bioinformatics and multiple machine learning algorithms, providing new ideas for studying the mechanism and treatment of age-related cardiac decline.

Keywords: aging; bioinformatics; heart failure; immune infiltration analysis; machine learning.

Figure 1

The study flowchart.

Figure 2

Identification of aging-related genes. (A) Gene and trait clustering dendrograms of HF. Gene clustering trees (dendrograms) obtained by hierarchical clustering of neighbor-based differences. (B) 9 gene co-expression modules of HF. The numbers in each cell means the correlation coefficient and p-value. (C) 57 genes promote both aging and HF. (D) 195 genes inhibit both aging and HF.

Figure 3

Function enrichment analysis of 252 aging-related genes. (A) GO enrichment analysis results. (B) KEGG enrichment analysis results.

Figure 4

Machine learning in screening key aging genes for HF. (A) Screening of key aging genes using the Lasso Model in up-regulated genes. The Lasso coefficient profiles were utilized to identify the optimal feature genes, with the optimal lambda determined by minimizing the partial likelihood deviance. Each coefficient curve in the left picture represents an individual gene. The solid vertical lines in the right picture represent the partial likelihood deviance, and the number of genes (n = 17) corresponding to the lowest point of the curve was deemed most suitable for the Lasso model. (B) Screening of key aging genes using the RF Model in up-regulated genes. The relative importance of overlapping candidate genes was calculated using the random forest approach. We present the results for the top 20 genes. (C) Screening of key aging genes using the SVM-RFE Model in up-regulated genes. The SVM-RFE algorithm was employed to further identify the optimal feature genes, based on the highest accuracy and lowest error obtained from the curves. The x-axis indicates the number of feature selections, while the y-axis represents the prediction accuracy. (D) Venn diagram illustrating the identification of 10 candidate genes for up-regulated genes through the aforementioned three algorithms. (E) Screening of key aging genes using the Lasso Model in down-regulated genes. (F) Screening of key aging genes using the RF Model in down-regulated genes. (G) Screening of key aging genes using the SVM-RFE Model in down-regulated genes. (H) Venn diagram shows that 4 key aging genes for down-regulated genes are identified via the above three algorithms.

Figure 5

The accuracy of top 50 machine- learning algorithm combinations.

Figure 6

Analysis of Immune Cell Infiltration. (A) Visualization from bar graphs of the proportions of 22 types of immune cells in elderly healthy samples, elderly heart failure samples, and young heart failure samples from GSE57338. (B) Expression of 2 dysregulated immune cells in elderly heart failure samples and elderly healthy samples (*indicates p<0.05, the same below). (C) Expression of 3 dysregulated immune cells in elderly heart failure samples and young heart failure samples. (D) Visualization of bar graphs showing the proportions of 22 types of immune cells at the circulating level in elderly heart failure samples and young heart failure samples from GSE77343. (E) The proportion of 1 dysregulated immune cell in elderly heart failure samples and young heart failure samples. This indicates differences between myocardial immune cell levels and circulating immune cell levels.