Integrated Gene Expression Profiling Analysis Reveals Potential Molecular Mechanisms and Candidate Biomarkers for Early Risk Stratification and Prediction of STEMI and Post-STEMI Heart Failure Patients

Abstract

Objective: To explore the molecular mechanism and search for the candidate differentially expressed genes (DEGs) with the predictive and prognostic potentiality that is detectable in the whole blood of patients with ST-segment elevation (STEMI) and those with post-STEMI HF.Methods: In this study, we downloaded GSE60993, GSE61144, GSE66360, and GSE59867 datasets from the NCBI-GEO database. DEGs of the datasets were investigated using R. Gene ontology (GO) and pathway enrichment were performed via ClueGO, CluePedia, and DAVID database. A protein interaction network was constructed via STRING. Enriched hub genes were analyzed by Cytoscape software. The least absolute shrinkage and selection operator (LASSO) logistic regression algorithm and receiver operating characteristics analyses were performed to build machine learning models for predicting STEMI. Hub genes for further validated in patients with post-STEMI HF from GSE59867.Results: We identified 90 upregulated DEGs and nine downregulated DEGs convergence in the three datasets (|log2FC| ≥ 0.8 and adjusted p < 0.05). They were mainly enriched in GO terms relating to cytokine secretion, pattern recognition receptors signaling pathway, and immune cells activation. A cluster of eight genes including ITGAM, CLEC4D, SLC2A3, BST1, MCEMP1, PLAUR, GPR97, and MMP25 was found to be significant. A machine learning model built by SLC2A3, CLEC4D, GPR97, PLAUR, and BST1 exerted great value for STEMI prediction. Besides, ITGAM and BST1 might be candidate prognostic DEGs for post-STEMI HF.Conclusions: We reanalyzed the integrated transcriptomic signature of patients with STEMI showing predictive potentiality and revealed new insights and specific prospective DEGs for STEMI risk stratification and HF development.

Screening of Lipid Metabolism-Related Gene Diagnostic Signature for Patients With Dilated Cardiomyopathy

“…A binomial distribution variable was then used in the LASSO classification, as well as the 1 standard error of the minimum criteria (the 1-SE criteria) lambda value used to build the model with good performance but the least number of variables for 10-fold cross-validation. AUC values were used to evaluate the model's ability to differentiate between DCM and normal samples by the pROC package in R (13,14). The data analysis process is depicted in Figure 1.…”

Machine learning-based integration develops biomarkers initial the crosstalk between inflammation and immune in acute myocardial infarction patients

“…This is consistent with the results we obtained in the ssGSEA single-gene association test. While NFIL3 ( 38 ) and MCEMP1 ( 39 ) currently with only a few omics studies demonstrated their potential relationship with AMI, our study points to the potential clinical value of both, which may be a viable direction for future research. It is worth noting that the expression levels of these key genes were verified by qRT-PCR, and the results were consistent with the results of bioinformatics.…”

A novel machine learning-derived four-gene signature predicts STEMI and post-STEMI heart failure

“…A recent study proposed the diagnostic value of a combination of five genes identified by logistic LASSO regression to discriminate STEMI patients from controls and nominated two genes as predictive biomarkers of post-STEMI HF. However, the robustness of this conclusion could be challenged by the heterogeneity of the sample source [ 25 ]. Another three-gene model for recognizing post-acute MI HF [ 30 ] suffered from data reuse and lack of validation, significantly impacting the generalizability.…”