What is the clinical evidence from this study?

Study design: Observational. Population: ST-segment elevation myocardial infarction (STEMI) and post-STEMI heart failure. Intervention: 5-gene machine learning prediction model (SLC2A3, CLEC4D, GPR97, PLAUR, BST1) vs. Healthy controls. Primary outcome: Prediction of STEMI (combination set) (AUC 0.9441).

December 10, 2020Open Access

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

Key Result

A machine learning model built with five hub genes (SLC2A3, CLEC4D, GPR97, PLAUR, and BST1) predicted STEMI with an AUC of 0.9441, while ITGAM and BST1 emerged as prognostic biomarkers for post-STEMI HF.

Study Design

Type

Observational

PICO

Population

ST-segment elevation myocardial infarction (STEMI) and post-STEMI heart failure

Intervention / Comparator

5-gene machine learning prediction model (SLC2A3, CLEC4D, GPR97, PLAUR, BST1) vs Healthy controls

Primary Outcome

Prediction of STEMI (combination set) — AUC 0.9441

Main Result

Effect estimate: AUC 0.9441

Limitations

The post-STEMI HF cohort is relatively small.
The databases of gene ontology will be revised such that the analysis may have to be repeated.
The results from the gene array and bioinformatic analysis require further biological proof-of-concept studies to verify.
Large-scale and prospective investigations are required to confirm the clinical feasibility of the proposed biomarkers.

Abstract

Abstract Objective To explore the molecular mechanism and search for the candidate biomarkers with predictive and prognostic potentiality that detectable in the whole blood of STEMI patients and post-STEMI HF patients. Methods In this study, we downloaded GSE60993, GSE61144, GSE66360, and GSE59867 datasets from the NCBI-GEO database. Differentially expressed genes (DEGs) of the datasets were investigated using R. Gene ontology and pathway enrichment were performed via ClueGO, CluePedia, and DAVID database. Protein interaction network was constructed via STRING. Enriched hub genes were analyzed by Cytoscape software. LASSO logistic regression algorithm and ROC analysis were performed to build machine learning models for predicting STEMI. Hub genes for further validated in post-STEMI HF patients from GSE59867. Results We identified 90 up-regulated DEGs and 9 down-regulated DEGs convergence in the three datasets (|log 2 FC| ≥ 0.8 and adjusted p value < 0.05). They were mainly enriched in Gene Ontology terms relating to cytokine secretion, pattern recognition receptors signaling pathway, and immune cells activation. A cluster of 8 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 biomarkers for post-STEMI HF. Conclusions We re-analyzed the integrated transcriptomic signature of STEMI patients showing predictive potentiality and revealed new insights and specific prospective biomarkers for STEMI risk stratification and HF development.

Bookmark

View Full Paper

Bookmark

View Full Paper

Cite This Study

Xu et al. (Thu,) conducted a observational in ST-segment elevation myocardial infarction (STEMI) and post-STEMI heart failure. 5-gene machine learning prediction model (SLC2A3, CLEC4D, GPR97, PLAUR, BST1) vs. Healthy controls was evaluated on Prediction of STEMI (combination set) (AUC 0.9441). A machine learning model built with five hub genes (SLC2A3, CLEC4D, GPR97, PLAUR, and BST1) predicted STEMI with an AUC of 0.9441, while ITGAM and BST1 emerged as prognostic biomarkers for post-STEMI HF.

synapsesocial.com/papers/6a10db07cfa01e990d9fac3d https://doi.org/https://doi.org/10.21203/rs.3.rs-118025/v1

Also Consider

Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: