Abstract

Background: Acute myocardial infarction (MI) is a serious emergency disease with high mortality. Hypoxia is associated with unfavorable outcomes in cancer patients. Nevertheless, there remains a shortage of effective hypoxia-related biomarkers to forecast the prognosis of acute MI patients and to identify targeted therapies.

Methods: First, data on acute MI patient samples and hypoxia-related genes were obtained based on public databases. Hypoxia-related gene scores were calculated by single sample Gene Set Enrichment Analysis (ssGSEA). Hypoxia-related hub genes in acute MI were screened via weighted correlation network analysis (WGCNA). Acute MI samples were analyzed for differentially expressed genes (DEGs) using the limma package and intersected with hub gene for hypoxia-related DEGs. Then, machine learning methods were used to identify hypoxia-related biomarkers in acute MI. Gene set enrichment analysis (GSEA) and immune infiltration analysis were performed on biomarkers. Targeted drug prediction and molecular docking were conducted based on biomarkers.

Results: The hypoxia-related gene score of the acute MI group was higher than the control group, and 319 hypoxia-related hub genes in acute MI were acquired. A total of 7 hypoxia-related DEGs were obtained by WGCNA and DEGs analysis. Then, 2 hypoxia-related biomarkers in acute MI, HAUS3 and SLC2A3, were identified based on machine learning algorithms. Both HAUS3 and SLC2A3 were enriched in the ribosome and spliceosome pathways. The expression levels of SLC2A3 and HAUS3 were correlated with immune cell infiltration. Furthermore, 8-hydroxyquinoline, perhexiline, and sotalol were selected as the targeted drugs, which could bind to HAUS3 and SLC2A3.

Conclusion: In short, we screened two important hypoxia-related biomarkers and three potential target drugs based on bioinformatics techniques. This provides new ideas and potential drug targets for early diagnosis and targeted therapy of acute MI.

Keywords: Acute myocardial infarction, hypoxia, biomarker, machine learning, drug discovery, molecular docking.