Bioinformatic identification and analysis of immune-related chromatin regulatory genes as potential biomarkers in idiopathic pulmonary fibrosis

BACKGROUND: Idiopathic pulmonary fibrosis (IPF) is a disease with a very poor prognosis. The search for new IPF biomarkers is particularly urgent due to the uncertainty of the mechanisms and treatment. Studies have shown that chromatin regulators (CRs) are involved in the development of IPF and are associated with tumor immunity. However, there are no studies on immune-related CRs in IPF. Therefore, we conducted a systematic study to analyze the expression levels and immune correlation of CRs in IPF tissues and normal tissues and to explore their potential as diagnostic biomarkers. METHODS: GSE53845, GSE179781 and GSE24206 datasets from Gene Expression Omnibus (GEO) database were merged into an integrated dataset as the training set; GSE70866 was used as the validation dataset. The cr-related differentially expressed genes (DEGs) between normal and IPF tissues were identified using the “Limma” software package. Weighted gene co-expression network analysis (WGCNA) was performed using the “WGCNA” package to screen eigengenes, which were intersected with DEGs to identify hub genes. The “ggcorrplot” package was used to analyze the correlation between hub genes and immunity, and immune-related hub genes were defined as immHub. A logistic regression model was constructed using immHub as the independent variable and whether the diagnosis was IPF as the dependent variable. RESULTS: One hundred and sixty-nine DEGs were identified between IPF and normal tissues. wGCNA identified 3 key modules in brown, green and yellow genes that were present in all 3 modules and met module membership (MM) >0.8 and gene significance (GS) >0.5 were called signature genes (n=390). Four intersecting genes were obtained by intersecting DEGs with signature genes (PADI4, IGFBP7, GADD45A, and SETBP1) all associated with immunity were defined as immHub genes Logistic regression models were constructed based on immHub genes. The area under the curve (AUC) of the ROC curve is used to evaluate the diagnostic accuracy of the logistic regression model for IPF. The AUC in the ROC analysis was 0.771 for the training dataset, and 0.759 for the validation dataset. CONCLUSIONS: PADI4, IGFBP7 and GADD45A may be biomarkers for IPF, which will provide assistance in the diagnosis, treatment and prognostic assessment of IPF patients, and provide an important basis for future studies on the relationship between CRs genes and IPF.