In Frontiers in cardiovascular medicine
BACKGROUND : Molecular biomarkers are widely used for disease diagnosis and exploration of pathogenesis. Pulmonary arterial hypertension (PAH) is a rapidly progressive cardiopulmonary disease with delayed diagnosis. Studies were limited regarding molecular biomarkers correlated with PAH from a broad perspective.
METHODS : Two independent microarray cohorts comprising 73 PAH samples and 36 normal samples were enrolled in this study. The weighted gene co-expression network analysis (WGCNA) was performed to identify the key modules associated with PAH. The LASSO algorithm was employed to fit a diagnostic model. The latent biology mechanisms and immune landscape were further revealed via bioinformatics tools.
RESULTS : The WGCNA approach ultimately identified two key modules significantly associated with PAH. For genes within the two models, differential expression analysis between PAH and normal samples further determined nine key genes. With the expression profiles of these nine genes, we initially developed a PAH diagnostic signature (PDS) consisting of LRRN4, PI15, BICC1, PDE1A, TSHZ2, HMCN1, COL14A1, CCDC80, and ABCB1 in GSE117261 and then validated this signature in GSE113439. The ROC analysis demonstrated outstanding AUCs with 0.948 and 0.945 in two cohorts, respectively. Besides, patients with high PDS scores enriched plenty of Th17 cells and neutrophils, while patients with low PDS scores were dramatically related to mast cells and B cells.
CONCLUSION : Our study established a robust and promising signature PDS for diagnosing PAH, with key genes, novel pathways, and immune landscape offering new perspectives for exploring the molecular mechanisms and potential therapeutic targets of PAH.
Duo Mengjie, Liu Zaoqu, Zhang Yuyuan, Li Pengfei, Weng Siyuan, Xu Hui, Wang Yu, Jiang Tianci, Wu Ruhao, Cheng Zhe
2022
diagnostic model, functional analysis, immune infiltration, machine learning, pulmonary arterial hypertension, weighted gene co-expression network analysis
