A neural network model was constructed by screening the potential biomarkers of aortic dissection based on genes associated with pyroptosis.

BACKGROUND

Aortic dissection (AD) is one of the crucial and common cardiovascular diseases, and pyroptosis is a novel cell delivery mechanism that is probably involved in the pathogenesis of various cardiovascular diseases. However, no study has investigated the role of pyroptosis in AD.

METHODS

We obtained two AD datasets, GSE153434 and GSE190635, from the Gene Expression Omnibus database. The differential expression of AD-related genes was determined by differential analysis, and their enrichment analysis was performed using Gene Ontology and Kyoto Encyclopedia of Genes and Genomes databases. Additionally, a protein-protein interaction network was established. Next, potential biomarkers were screened by Lasso regression analysis, and a neural network model was constructed. Finally, the potential biomarkers were validated by constructing a mouse model of AD.

RESULTS

A total of 1033 differentially expressed related genes were distinguished and these genes were mainly associated with the phosphoinositide 3-kinase (PI3K)/protein kinase B (Akt) and mitogen-activated protein kinase signaling pathways. The Lasso regression results showed five potential biomarkers, namely platelet endothelial cell adhesion molecule-1 (PECAM1), caspase 4 (CASP4), mixed lineage kinase domain-like pseudokinase (MLKL), APAF1-interacting protein (APIP), and histone deacetylase 6 (HDAC6) and successfully constructed a neural network model to predict AD occurrence. The results showed that CASP4 and MLKL were highly expressed, whereas PECAM1 and HDAC6 were lowly expressed in AD samples, and no statistically significant difference was observed in APIP expression in AD samples.

CONCLUSION

Pyroptosis plays a crucial role in AD occurrence and development. Moreover, the five potential biomarkers identified in the present study can act as targets for the early diagnosis of AD in patients.