Background: Acute myocardial infarction (AMI) is a leading cause of morbidity and mortality worldwide. Circulating endothelial cells (CECs) have been reported to be involved with the early stages of AMI. The specific objective of our study was to discover early diagnostic markers of CECs in circulation using bioinformatics analysis.
Methods: Raw microarray data of the GSE66360 dataset were acquired from the Gene Expression Omnibus (GEO) database. The R software was used to filtrate differentially expressed genes (DEGs) from the discovery cohort of GSE66360 (n = 43). A weighted gene co-expression network analysis (WGCNA) was performed to explore the key modules connected with AMI. Next, main roles of the pathological states in AMI were analyzed using GO and KEGG and PPI networks. Diagnostic biomarkers were selected and identified using three machine learning algorithms. Additionally, the expression and diagnostic efficiency of hub genes were verified in the validation cohort (n = 56).
Results: 366 DEGs were identified (20 upregulated and 306 downregulated). A total of 276 intersecting genes were markedly associated with AMI in the pink and turquoise modules. Based on multiple machine learning algorithms and independent validation, six genes including LILRA1, CCL20, IL1R2, TYROBP, CXCL16 and NFKBIA were identified as hub genes and showed satisfactory diagnostic efficiency both in the discovery cohort and validation cohort.
Conclusion: Our data provides evidence supporting a list of six hub genes to be trapped in the pathophysiology of AMI and proposes them as candidate biomarkers for the early diagnosis of AMI.
Keywords: Acute myocardial infarction; Circulating endothelial cells; Hub genes; Machine learning; WGCNA.
© 2025 The Authors. Published by Elsevier Inc.