Construction of a neural network diagnostic model for idiopathic pulmonary fibrosis and investigation of immune infiltration characteristics

Background: Idiopathic Pulmonary Fibrosis (IPF) is a fatal interstitial lung disease characterized by rapid progression and unpredictable clinical course. However, there are currently limited tools available for the diagnosis and treatment of IPF. This study aims to identify and screen potential biomarkers for IPF diagnosis, thereby providing new diagnostic approaches. Methods: We selected datasets from GEO database and screened for differentially expressed genes (DEGs) between IPF and normal tissues and identified IPF-related modules through WGCNA Subsequently, we employed two machine learning strategies to screen potential diagnostic biomarkers. Candidate biomarkers were quantitatively evaluated using ROC curves to identify key diagnostic genes, followed by the construction of a nomogram. Furthermore, the expression of these genes was validated using RT-qPCR. Next, we conducted immune infiltration analysis, GSEA, and targeted drug prediction. Finally, we developed an artificial neural network model for IPF. Results: We identified ASPN, COMP, and GPX8 as candidate genes for IPF biomarkers, with AUC values exceeding 0.90. These genes were validated by RT-qPCR. Immunoinfiltration analysis revealed that specific immune cell types are closely related to IPF, suggesting that these immune cells may play a significant role in the pathogenesis of IPF. Conclusion: ASPN, COMP, and GPX8 have been identified as potential diagnostic genes for IPF, and the most relevant immune cell types have been determined.