Integrated Bioinformatics Analysis and Machine Learning to Identify Biomarkers and Fibroblast-related Features of Idiopathic Pulmonary Fibrosis

BACKGROUND: Idiopathic Pulmonary Fibrosis (IPF) is an interstitial lung disease with an undefined etiology and poor prognosis. However, therapeutic options for IPF are extremely limited at present. Therefore, it is particularly important to explore its biomarkers and therapeutic targets. METHODS: In this study, we integrated multi-omics analysis and Summary-data-based Mendelian Randomization (SMR) analysis to identify the causative genes of IPF associated with iron metabolism and explore the related characteristics of fibroblasts. Multiple machine learning algorithms were used to screen diagnostic biomarkers and build diagnostic models related to iron metabolism. RESULTS: The results of our analysis indicate that there is a close correlation between the STEAP2 gene and IPF. In addition, this study revealed a stronger interaction between STEAP2-positive fibroblasts and endothelial cells and identified the differentiation trajectory of fibroblast subpopulations. The diagnostic model based on the three biomarkers STEAP2, SFRP2, and CCL13 demonstrated promising diagnostic efficacy for IPF, with the XGBoost model achieving the highest diagnostic accuracy. DISCUSSION: The results of Mendelian randomization and multi-omics analyses revealed the important role of STEAP2 in IPF. Furthermore, the diagnostic model screened for iron metabolism-related genes and validated their diagnostic value in IPF. CONCLUSION: We identified that STEAP2 plays an important regulatory role in IPF and developed a diagnostic model based on characteristic biomarkers. Overall, this research provides new insights into the pathogenesis of IPF from both biological and genetic perspectives, offering potential therapeutic targets for its treatment.