Abstract: BACKGROUND: Mitochondria are of great significance in the injury and repair of acute myocardial infarction, so it is of great clinical significance to explore the pathogenesis and progression of acute myocardial infarction based on mitochondrial genes. 
OBJECTIVE: To explore whether mitochondrial genes can be used as reliable biomarkers to assess the progression of acute myocardial infarction. 
METHODS: The acute myocardial infarction dataset was downloaded from the Gene Expression Omnibus GSE66360 and GSE12288, and the human mitochondrial gene set was obtained from the mitochondrial protein database. Differential gene analysis and weighted correlation network analysis were performed on the acute myocardial infarction dataset GSE66360, and the genes were obtained for protein-protein interaction analysis, gene ontology, and kyoto encyclopedia of genes genomes analysis. The intersection genes were intersected with human mitochondrial genes to obtain differential mitochondrial genes. Gene set enrichment analysis and immune infiltration analysis were performed on differential mitochondrial genes. Receiver operating characteristic curve analysis was performed in the external dataset GSE12288 to verify the expression characteristics of the differential mitochondrial genes. 
RESULTS AND CONCLUSION: (1) A total of 548 differential genes were obtained for acute myocardial infarction. Differential gene analysis and weighted gene co-expression network analysis showed that the Blue module contained the most genes (4 992). There were 116 intersecting genes, among which tumor necrosis factor, interleukin-1B, interleukin-6, Toll-like receptor 4, and interleukin-10 were the core genes. (2) Gene ontology analysis showed that the biological process mainly involved inflammatory response, positive regulation of tumor necrosis factor production, cell surface receptor signaling pathway. Cell composition mainly involved tertiary granular membrane, plasma membrane outer layer, plasma membrane, etc. Molecular function mainly involved immunoglobulin G binding, transmembrane signal receptor activity, chemokine activity, etc. Kyoto encyclopedia of genes genomes analysis showed that these genes were mainly involved in tumor necrosis factor, interleukin-17, and nuclear factor kappa-light-chain-enhancer of activated B cell pathways. (3) A total of eight differential mitochondrial genes were obtained, and four characteristic genes were screened after least absolute shrinkage and selection operator and proportional hazards model analysis, including phorbol-12-myristate-13-acetate-induced protein1, BCL2 related protein A1, solute carrier family 25 member 37, and deoxyribonucleic acid polymerase beta, and corresponded to tumor protein 53, oxidative phosphorylation, sulfur metabolism, and glycerol phospholipid metabolism pathways, respectively. (4) Receiver operating characteristic curve analysis showed that the four characteristic genes were all of diagnostic significance, and were negatively correlated with resting dendritic cells and naïve B cells. (5) These results suggest that the expression characteristics of phorbol-12-myristate-13-acetate-induced protein1, BCL2 related protein A1, solute carrier family 25 member 37, and deoxyribonucleic acid polymerase beta can be used as potential biomarkers to predict mitochondrial function in acute myocardial infarction, which can further improve the accuracy of prediction of acute myocardial infarction.

Key words: ">mitochondria, acute myocardial infarction, bioinformatics, machine learning, immune infiltration analysis, oxidative stress, inflammation, apoptosis