AUTHOR=Yang Yujia , Cai Yue , Zhang Yuan , Yi Xu , Xu Zhiqiang 

TITLE=Identification of Molecular Subtypes and Key Genes of Atherosclerosis Through Gene Expression Profiles

JOURNAL=Frontiers in Molecular Biosciences

VOLUME=Volume 8 - 2021

YEAR=2021

URL=https://www.frontiersin.org/journals/molecular-biosciences/articles/10.3389/fmolb.2021.628546

DOI=10.3389/fmolb.2021.628546

ISSN=2296-889X

ABSTRACT=Atherosclerotic cardiovascular disease (ASCVD) caused by atherosclerosis (AS) is one of the highest causes of mortality in the worldwide. Although there have been many studies on AS, its etiology remains unclear. In order to carry out molecular characterization of different types of AS, we retrieved two datasets composed of 151 AS samples and 32 normal samples from the Gene Expression Omnibus database. Using the NMF algorithm, we successfully divided the 151 AS samples into two subgroups. We then compared the molecular characteristics between the two groups using weighted gene co‐expression analysis (WGCNA) analyses and identified six key modules associated with the two subgroups. KEGG and GO enrichment analysis were used to identify the potential functions and pathways associated with the modules. In addition, we used the cytoscape software to construct and visualize protein-protein networks so as to identify key genes in the modules of interest. Three hub genes including PTGER3, GNAI1 and IGFBP5 were further screened using the least absolute shrinkage and selection operator (LASSO) and support vector machine-recursive feature elimination (SVM-RFE) algorithms. Since the modules were associated with immune pathways, we performed immune cell infiltration analysis. We discovered that a significant difference in the level of immune cell infiltrationby naïve B cells, CD8 T cells, T regulatory cells (Tregs), resting NK cells, Monocytes, Macrophages M0, Macrophages M1 and Macrophages M2  between the two subgroups. In addition, we observed the three hub genes were positively correlated with Tregs, but negatively correlated with Macrophages M0. We also found that the three key genes are differentially expressed between normal and diseased tissue, as well as in the different subgroups. ROC results showed a good performance in the validation dataset. These results may provide novel insight into cellular and molecular characteristics of AS and potential markers for diagnosis and targeted therapy