Unlocking potential biomarkers bridging coronary atherosclerosis and pyrimidine metabolism-associated genes through an integrated bioinformatics and machine learning approach

Abstract

Background: This study delves into the intricate landscape of atherosclerosis (AS), a chronic inflammatory disorder with significant implications for cardiovascular health. AS poses a considerable burden on global healthcare systems, elevating both mortality and morbidity rates. The pathological underpinnings of AS involve a marked metabolic disequilibrium, particularly within pyrimidine metabolism (PyM), a crucial enzymatic network central to nucleotide synthesis and degradation. While the therapeutic relevance of pyrimidine metabolism in diverse diseases is acknowledged, the explicit role of pyrimidine metabolism genes (PyMGs) in the context of AS remains elusive. Utilizing bioinformatics methodologies, this investigation aims to reveal and substantiate PyMGs intricately linked with AS.

Methods: A set of 41 candidate PyMGs was scrutinized through differential expression analysis. GSEA and GSVA were employed to illuminate potential biological pathways and functions associated with the identified PyMGs. Simultaneously, Lasso regression and SVM-RFE were utilized to distill core genes and assess the diagnostic potential of four quintessential PyMGs (CMPK1, CMPK2, NT5C2, RRM1) in discriminating AS. The relationship between key PyMGs and clinical presentations was also explored. Validation of the expression levels of the four PyMGs was performed using the GSE43292 and GSE9820 datasets.

Results: This investigation identified four PyMGs, with NT5C2 and RRM1 emerging as key players, intricately linked to AS pathogenesis. Functional analysis underscored their critical involvement in metabolic processes, including pyrimidine-containing compound metabolism and nucleotide biosynthesis. Diagnostic evaluation of these PyMGs in distinguishing AS showcased promising results.

Conclusion: In conclusion, this exploration has illuminated a constellation of four PyMGs with a potential nexus to AS pathogenesis. These findings unveil emerging biomarkers, paving the way for novel approaches to disease monitoring and progression, and providing new avenues for therapeutic intervention in the realm of atherosclerosis.

Keywords: Atherosclerosis (AS); Bioinformatics; Lasso regression; Pyrimidine Metabolism Genes (PyMGs); SVM-RFE.

Fig. 1

Framework

Fig. 2

Principal Component Analysis. a Analysis of difference. b Analysis of correlation

Fig. 3

For PyMGs, GO, and KEGG analyses were performed. a The GO circle illustrates the scatter map of the selected gene's logFC. b The KEGG barplot and bubble illustrates the scatter map of the logFC of the indicated gene

Fig. 4

The development of the PyMGs signature. a Regression of the 4 AS-related genes using LASSO. b Cross-validation is used in the LASSO regression to fine-tune parameter selection. c-d Accuracy and error of this model. e Venn. f AUC of 4 hub genes. g AUC of train group

Fig. 5

GSEA of Analysis in NT5C2 and RRM1. a GO. b KEGG

Fig. 6

Expression of Immune cells. a Expression of immune cells in different clusters. b Correlation between PyMGs and immune cells

Fig. 7

GSVA of Analysis in NT5C2 and RRM1. a GO. b KEGG

Fig. 8

Drug-gene interactions. Note: Red circles are up-regulated genes, green hexagons are down-regulated genes, and blue squares are associated drugs

Fig. 9

miRNAs-LncRNAs shared Genes Network. Note: Red circles are mrnas, blue quadrangles are miRNAs, and green triangles are lncRNAs

Fig. 10

Hub gene and Model verification. a-b Hub genes were validated. c-d Residual expression patterns. e AUC of model. f AUC of GSE9820 group