PLA2G2A is a candidate biomarker for sepsis diagnosis in mice through machine learning algorithms

Abstract

In this study, machine-learning screening was implemented on human transcriptomic data from public GEO datasets and independent experiments were served to validate candidate biomarkers for sepsis in mice models. Microarray data were obtained from the GSE236713 dataset, and differentially expressed genes (DEGs) were screened. We carried out enrichment analysis with the help of gene ontology (GO) and Kyoto encyclopedia of genes and genomes (KEGG). We utilized machine learning algorithms, such as least absolute shrinkage and selection operator (LASSO), logistic regression (LR), support vector machine (SVM). For the key genes, we conducted single-gene enrichment and immune Infiltration Analysis. We selected C57BL/6J SPF-grade male mice to construct a sepsis model. The proteins were then detected via western blotting (WB), tissue homogenization, tissue staining, immunohistochemistry (IHC), and cytokines. Multiple groups of DEGs were identified via dataset analysis. Weighted gene co-expression network analysis (WGCNA) was used to identify related modules. Machine learning was used to screen key genes, such as cholinergic receptor muscarinic 1 (CHRM1) and Secreted phospholipase A2 (sPLA2-IIA) (encoded by PLA2G2A ). sPLA2-IIA expression was also higher in the mouse sepsis model group than the sham group. And its expression was implying regulatory effects. The successful mouse model is characterized by elevated levels of IL-6, TNF-α, and multi-organ damage. In this study, we identified PLA2G2A as the candidate biomarker for diagnosis of sepsis in mice models through machine learning algorithms. And the data obtained may be helpful for subsequent research on sepsis pathogenesis.