Frontiers in Cardiovascular Medicine (Apr 2022)

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

  • Man Xu,
  • Man Xu,
  • Ying-ying Guo,
  • Ying-ying Guo,
  • Dan Li,
  • Dan Li,
  • Xian-feng Cen,
  • Xian-feng Cen,
  • Hong-liang Qiu,
  • Hong-liang Qiu,
  • Yu-lan Ma,
  • Yu-lan Ma,
  • Si-hui Huang,
  • Si-hui Huang,
  • Qi-zhu Tang,
  • Qi-zhu Tang

DOI
https://doi.org/10.3389/fcvm.2022.853468
Journal volume & issue
Vol. 9

Abstract

Read online

BackgroundDilated cardiomyopathy (DCM) is characterized by enlarged ventricular dimensions and systolic dysfunction and poor prognosis. Myocardial lipid metabolism appears abnormal in DCM. However, the mechanism of lipid metabolism disorders in DCM remains unclear.MethodsA gene set variation analysis (GSVA) were performed to estimate pathway activity related to DCM progression. Three datasets and clinical data downloaded from the Gene Expression Omnibus (GEO), including dilated cardiomyopathy and donor hearts, were integrated to obtain gene expression profiles and identify differentially expressed genes related to lipid metabolism. GO enrichment analyses of differentially expressed lipid metabolism-related genes (DELs) were performed. The clinical information used in this study were obtained from GSE21610 dataset. Data from the EGAS00001003263 were used for external validation and our hospital samples were also tested the expression levels of these genes through RT-PCR. Subsequently, logistic regression model with the LASSO method for DCM prediction was established basing on the 7 DELs.ResultsGSVA analysis showed that the fatty acid metabolism was closely related to DCM progression. The integrated dataset identified 19 DELs, including 8 up-regulated and 11 down-regulated genes. A total of 7 DELs were identified by further external validation of the data from the EGAS00001003263 and verified by RT-PCR. By using the LASSO model, 6 genes, including CYP2J2, FGF1, ETNPPL, PLIN2, LPCAT3, and DGKG, were identified to construct a logistic regression model. The area under curve (AUC) values over 0.8 suggested the good performance of the model.ConclusionIntegrated bioinformatic analysis of gene expression in DCM and the effective logistic regression model construct in our study may contribute to the early diagnosis and prevention of DCM in people with high risk of the disease.

Keywords